Visual decision benchmark questions and images

200 questions from the October 15, 2026 benchmark. The public gallery includes 221 tested image files; six charts are referenced but withheld pending third-party terms review. Gold answers remain private. Open this file locally after cloning the repository to browse the images.

VB0001 Connectivity and reachability

In the diagram each arrow runs from one numbered circle to another. Following arrows only in the direction they point, what is the smallest number of arrows needed to get from node 1 to node 6?

An arrow ends where its arrowhead touches a circle. Where two lines cross in open space without ending at a circle, they are not connected. Moving against an arrowhead is not allowed.

CCA0005

Dimitris131, Public domain, via Wikimedia Commons (File:Directed graph example el.svg); credit: Own work, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0002 Connectivity and reachability

Starting at node h and following arrows only in the direction they point, how many of the other seven nodes can be reached? (Do not count h itself.)

Only arrows count. The shaded regions and dotted outlines are decoration and do not connect or separate anything. A pair of curved arrows between two nodes means each can reach the other.

CCA0006

Д.Ильин: vectorization, CC0, via Wikimedia Commons (File:Scc-1.svg); credit: File:Scc.png by ℚuackor, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0003 Connectivity and reachability

Following the arrows from the box labelled a:main, how many of the other nine boxes can be reached?

An arrow points from a caller to the box it calls. A box that has an arrow pointing out of it does not become reachable from a:main for that reason alone. A looping arrow that returns to its own box adds nothing.

CCA0023

Dmccreary, CC BY-SA 3.0, via Wikimedia Commons (File:Sample-single-module-call-graph.png); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0004 Connectivity and reachability

In this call graph, which box is the end point of arrows that come from two different boxes?

Numbers beside the arrows are call counts and do not change which arrows exist. Count only arrowheads that touch the box.

CCA0024

GeraldKaszuba, CC BY-SA 3.0, via Wikimedia Commons (File:A Call Graph generated by pycallgraph.png); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0005 Connectivity and reachability

Following arrows only in the direction they point, what is the largest number of arrows on a path from tgmath.h to nlib.h that never visits the same box twice?

An arrow ends at the box where its arrowhead touches. Two boxes can have arrows in both directions between them, but a path may not return to a box it has already visited.

CCA0025

Jobin RV, CC BY-SA 3.0, via Wikimedia Commons (File:Dependency Graph for tgmath.h.png); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0006 Connectivity and reachability

The dotted link between F and K is out of service (it carries a red X) and may not be used. Ignoring the bold highlighting, what is the smallest number of links on a route from F to Q?

Each line between two circles is one link. Count every link on the route. Lines meet only at circles.

CCA0026

Kjerish, CC BY-SA 4.0, via Wikimedia Commons (File:17 node network with a single link failure.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0007 Connectivity and reachability

Following arrow directions, how many different routes lead from the bottom circle labelled read_graph to the circle labelled count?

Rectangles and circles are both stops on a route. Two routes are different if they differ in at least one arrow. Arrows keep their direction and lines meet only where they end.

CCA0027

Daniee N H, CC BY-SA 4.0, via Wikimedia Commons (File:Example DASK.png); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0008 Connectivity and reachability

Starting at the circle labelled with the symbol for the caesium frequency (the top circle, written as a Greek letter followed by a subscript 'Cs') and following arrows only in the direction they point, how many other circles can be reached?

Arrows point from one circle to another; a circle with only outgoing arrows cannot be reached from elsewhere. Count each reached circle once.

CCA0028

Emilio Pisanty, CC BY-SA 4.0, via Wikimedia Commons (File:Unit relations in the new SI.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0009 Connectivity and reachability

The orange line starts at the box labelled 'Call A.add'. At which box does the arrowhead of that orange line point?

Lines may cross in open space without connecting. Each line keeps its colour from start to arrowhead.

CCA0031

Bautsch, CC0, via Wikimedia Commons (File:Zyklischer.Import.png); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0010 Connectivity and reachability

Starting at node G and following arrows only in the direction they point, how many other nodes can be reached?

Do not count G itself. The coloured rounded boxes and their letters (X, Y, Z, W, V) are decoration and do not connect or separate anything. Where two curved arrows run between the same two nodes in opposite directions, each node can reach the other.

CCA0139

RaspberryJamWikipedian, CC BY-SA 4.0, via Wikimedia Commons (File:Kosaraju stack after first pass.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0011 Navigation and planning

On this Bucharest metro diagram, at which station should a rider change lines to travel from Preciziei (west end of the red line) to Tudor Arghezi (south end of the blue line) with the fewest changes of line?

Use only solid line segments; ignore dotted or striped segments. A station drawn as two circles joined by a white link, or as one shared circle, is an interchange. Two stations that are merely close together without a link are not an interchange.

CCA0007

Mliu92, CC BY-SA 4.0, via Wikimedia Commons (File:Metro București.svg); credit: Own work, inspired by sources listed below with station names and rough routing from Harta reţelei de metrou din municipiul Bucureşti. Metrorex (April 2016). Archived from the original on 5 January 2017. Retrieved on 5 January 2016., CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0012 Navigation and planning

Using this Prague metro plan, what is the smallest number of station-to-station moves from Nemocnice Motol to Letňany?

Count one move between adjacent station symbols, including departures from and arrivals at the lettered square terminal markers. Changing lines at a shared station costs no moves. Ignore the dotted dark-blue line and the pale-blue river. Minimize moves, regardless of the number of line changes.

CP_CCA0022_v1

Bilykralik16, CC BY-SA 4.0, via Wikimedia Commons (File:Prague metro plan 2022.svg); credit: Own work with the base of File:Prague metro plan 2015.svg, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0013 Navigation and planning

What is the smallest total distance for a trip from Maldon to Dunwich?

Numbers beside the lines are distances. A route's length is the sum of the numbers on the lines it uses. Lines that cross in open space without a circle do not connect. A trip may pass through other towns. If two routes tie for shortest, the question is about the common length.

CP_CCA0078_v1

Pluke, CC0, via Wikimedia Commons (File:CPT-Graphs-undirected-weighted.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0014 Navigation and planning

What is the smallest total distance from node 6 to node 7?

Numbers beside the lines are distances. A route's length is the sum of the numbers on the lines it uses. Lines that cross in open space without a circle do not connect.

CCA0083

BagLuke, CC0, via Wikimedia Commons (File:Weighted friendship graph.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0015 Navigation and planning

To travel from Odivelas (north end of the yellow line) to Cais do Sodré (south end of the green line) with as few changes of line as possible, at which station do you change?

Use only solid line sections; dotted sections are under construction (see the legend) and cannot be used. A white circle marks a station where lines meet (common station). Choose the single station that allows the fewest changes.

CP_CCA0084_v1

Laura Sun, CC BY-SA 3.0, via Wikimedia Commons (File:Lisbon metro.svg); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0016 Navigation and planning

How many stops must a passenger travel from Tempelhof to reach Grenzallee, counting Grenzallee but not Tempelhof?

The line has one branch. A stop is any black square on the route. The route takes the branch to the right at the junction. Ignore the notes beside the open markers (they are not on this route).

CCA0087

Grummelbacke, CC0, via Wikimedia Commons (File:Nord-sued-u-bahn berlin karte 1930.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0017 Visual rule application

Apply the Game of Life rules to the pattern in the picture two times in a row. How many live cells are there after the second generation?

Black squares are live cells, white squares are dead cells, and every cell outside the picture is dead; the grid is unbounded in all directions. Rules: a live cell with 2 or 3 live neighbours (the 8 surrounding cells, diagonals included) stays alive, otherwise it dies; a dead cell with exactly 3 live neighbours becomes alive. All cells change together in each generation.

CCA0008

Sarang, Public domain, via Wikimedia Commons (File:Game of life 3x3 cell pento-R.svg); credit: Own work, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0018 Visual rule application

The picture is a Turing-machine state diagram. The machine starts in state A. The tape is infinite in both directions and every cell holds 0 except the single cell under the head, which holds 1. When the machine reaches the halting state H, how many cells on the tape hold 1?

An arrow label 'x/P,D' means: if the symbol under the head is x, write 1 (P), move the head one cell in direction D (L = left, R = right), then go to the state the arrow points to. The arrow from 'start' only marks the initial state. H is the halting state.

CCA0009

Diego Queiroz, CC BY-SA 3.0, via Wikimedia Commons (File:State diagram 3 state busy beaver 2B.svg); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0019 Visual rule application

The machine starts in state a (the state with the entering arrow). Which one of these input strings does it accept?

A string is read left to right, one symbol per arrow, following the arrow labelled with that symbol. It is accepted if, after the last symbol, the machine is in a state drawn with a double circle.

CCA0068

Pat Hawks, CC BY 4.0, via Wikimedia Commons (File:DFA to be minimized.svg); credit: Own work, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0020 Visual rule application

The machine starts in S0 (the state with no arrow into it). What output sequence does it produce for the input sequence 1, 0, 0, 1, 1?

An arrow labelled x/y means: on input x, output y and move along that arrow. Write the five outputs in order.

CCA0069

Jake Choules, Public domain, via Wikimedia Commons (File:Mealy machine state diagram.svg); credit: Own work, derived from Mealymachine_jaredwf.png., Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0021 Visual rule application

The machine starts in q0. For the input sequence z, y, y, x, x, what output string is produced?

Each state shows its output letter in its lower half. Write the output of the starting state first, then the output of each state reached after every input symbol. An arrow labelled with several letters (for example x, z) is followed for any of them.

CCA0071

Stiegenaufgang, based on NikolausMayer's work, CC BY-SA 4.0, via Wikimedia Commons (File:MooreMachineExample.svg); credit: Based on MooreAutomaton.png - vectorized, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0022 Visual rule application

The machine starts in state A (the top-left state). Which state is it in after reading the input 1, 1, 0, 1, 1, 1?

Follow the arrow labelled with each input symbol, one symbol at a time. A looping arrow keeps the machine in the same state.

CCA0072

ColinStocks, CC0, via Wikimedia Commons (File:ExampleMooreMachine.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0023 Visual rule application

The machine starts in the Locked state (the arrow from the black dot). For the inputs Push, Coin, Coin, Push, Push in that order, what is the output sequence?

An arrow labelled input/output means: on that input, produce that output and move along the arrow (a loop stays in the same state).

CCA0073

Chetvorno, CC0, via Wikimedia Commons (File:Turnstile finite-state machine with actions.svg); credit: Own work based on: Turnstile state machine colored.svg, with minor modifications, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0024 Visual rule application

The machine starts in S1 with its head on the leftmost symbol of the tape 1 1 0 1 1; every other cell holds 0. When it halts, how many cells on the tape hold 1?

An arrow labelled a/b,D means: if the symbol under the head is a, write b, move the head one cell in direction D (L = left, R = right), then go to the state the arrow points to. The machine halts when no arrow leaves the current state for the symbol under the head. The tape is unbounded in both directions.

CCA0074

Sn KGS, CC BY-SA 4.0, via Wikimedia Commons (File:Example of a Turing machine.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0025 Visual rule application

Starting from q0, which one of these strings is accepted by the machine?

q0 is the start state (the only state with no incoming arrow). A string is accepted if at least one path through the machine reads the whole string and ends in a state drawn with a double circle. A path that gets stuck (no arrow for the next symbol) does not count.

CCA0075

HeatSink, CC BY-SA 3.0, via Wikimedia Commons (File:Nfa several.svg); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0026 Visual rule application

Apply the Game of Life rules once to the pattern in the picture. How many live cells are there after that one generation?

Black squares are live cells, white squares are dead cells, and every cell outside the picture is dead; the grid is unbounded in all directions. A live cell with 2 or 3 live neighbours (the 8 surrounding cells, diagonals included) stays alive, otherwise it dies; a dead cell with exactly 3 live neighbours becomes alive. All cells change together.

CCA0076

Bryan.burgers, Public domain, via Wikimedia Commons (File:Game of life infinite1.svg); credit: Own work, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0027 Constraint satisfaction

Under the usual sudoku rules (no repeated digit in any row, column or 3x3 box), which digits could legally be written in the empty centre cell, row 5 column 5? Count rows from the top and columns from the left.

Treat every digit shown, black or green, as already placed. Heavy lines mark the 3x3 boxes; dashed lines are ordinary cell borders. Do not try to solve the whole puzzle; ask only which digits do not conflict with the placed digits.

CCA0015

Amahoney, CC BY-SA 4.0, via Wikimedia Commons (File:Sudoku partial 1.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0028 Constraint satisfaction

Fill the 6x6 grid with the digits 1 to 6 so that every row and every column contains each digit exactly once, and every heavily outlined group of cells produces the number in its corner using the operation shown. Exactly one completed grid satisfies every rule. Which choice is the bottom row of that grid, read left to right?

For + the digits in the group add up to the number; for x they multiply to the number; for - (two cells) the larger digit minus the smaller equals the number; for / (two cells) the larger digit divided by the smaller equals the number. Groups with no operation symbol shown do not occur; each group has exactly one label in its top-left cell. A digit may repeat inside a group only if its cells are in different rows and columns.

CCA0017

Underpants, SGBailey, Public domain, via Wikimedia Commons (File:KenKenProblem.svg); credit: SVG version of Kenkenproblem.png, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0029 Constraint satisfaction

Fill the 5x5 grid with the digits 1 to 5 so that no digit repeats in any row or column and every inequality sign is true. Which digit goes in the cell in row 5, column 3? (Count rows from the top and columns from the left.)

A sign between two neighbouring cells points to the smaller one: 'a > b' means the left cell is larger than the right cell, and 'a < b' means the left cell is smaller. The grid has exactly one solution.

CCA0090

User:Pruneau, CC BY-SA 3.0, via Wikimedia Commons (File:Futoshiki1.svg); credit: File:Futoshiki1.png, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0030 Constraint satisfaction

Solve this 6x6 Light Up puzzle. How many lamps does the completed puzzle contain?

Black squares are walls. A numbered black square must have exactly that many lamps in its four edge-adjacent squares. A lamp lights its own square and every white square in the same row and column up to the next black square. Every white square must be lit, and no lamp may be lit by another lamp. The puzzle has exactly one solution.

CCA0091

Life of Riley, Public domain, via Wikimedia Commons (File:Light up 1.svg); credit: Own work, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0031 Constraint satisfaction

Solve this 10x10 Light Up puzzle. How many lamps does the completed puzzle contain?

Black squares are walls. A numbered black square must have exactly that many lamps in its four edge-adjacent squares. A lamp lights its own square and every white square in the same row and column up to the next black square. Every white square must be lit, and no lamp may be lit by another lamp. The puzzle has exactly one solution.

CCA0092

Amit6, CC BY-SA 3.0, via Wikimedia Commons (File:Lightup.svg); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0032 Constraint satisfaction

Solve this Hitori puzzle. Which one of these cells must be shaded black in the unique solution? (Rows count from the top, columns from the left.)

Shade some cells black so that (a) no number appears more than once among the unshaded cells of any row or column, (b) no two shaded cells share an edge, and (c) all unshaded cells form a single group connected by edges. The puzzle has exactly one solution.

CCA0093

Own work, Dan1980, CC BY 1.0, via Wikimedia Commons (File:Hitori.svg); credit: Own work, based on File:Hitori.jpg, CC BY 1.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0033 Constraint satisfaction

Solve this Kakuro. Which number goes in the white cell in row 6, column 4 (bottom row, middle of the three white cells)?

Each white cell holds a digit from 1 to 9. A number in the upper right of a grey cell is the sum of the run of white cells to its right; a number in the lower left is the sum of the run of white cells below it. Digits may not repeat within a run. The 4 is already placed. The red circles are marks and carry no meaning. The puzzle has exactly one solution.

CCA0094

Samuli Lintula, Public domain, via Wikimedia Commons (File:Kakuro.01.svg); credit: Own work, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0034 Constraint satisfaction

Using the weekday table, which itinerary meets all four conditions: (1) leave Sausalito at 9:00 or later, (2) arrive in San Francisco no later than 11:00, (3) spend more than 2 hours in San Francisco between arriving and leaving on the return ferry, and (4) be back in Sausalito by 2:00 PM?

The two left columns give the Sausalito-to-San Francisco ferries (departure, then arrival in the same row). The two right columns give the San Francisco-to-Sausalito ferries (departure, then arrival in the same row). Times above the horizontal rule drawn across the table are AM and times below it are PM.

CCA0100

Piergiuliano Chesi, Public domain, via Wikimedia Commons (File:San Francisco Golden Gate Ferry timetable June 1974 02.jpg); credit: Own work from scan, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0035 Constraint satisfaction

Which weekday satisfies both rules: (1) two lessons in a row on that day are the same subject, and (2) the last lesson of the day (the 12:10 row) is Deutsch?

Read each day's column from top (7:55) to bottom (12:10). 'In a row' means two neighbouring rows in the same column.

CCA0101

unknown, Public domain, via Wikimedia Commons (File:Stundenplan.PNG); credit: own work with Shrek!, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0036 Flowcharts and decision trees

An item is in the inbox. It is actionable. It is not a single step, so a project is created. The first subtask of that project would take about 10 minutes, you must do it yourself, and it has to happen at a specific time. Following the flowchart, which box does that first subtask end in?

Treat the first subtask as re-entering the chart at the box the arrow labelled 'First subtask from the project' points to. Follow Yes/No labels exactly as drawn.

CCA0001

Flexagoon, CC0, via Wikimedia Commons (File:GTD flowchart.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0037 Flowcharts and decision trees

A sentence in an article has no citation. You do not believe it is false. You find that sources do say it, and you could add a citation but you do not want to help out. Following the flowchart from START, which end box do you reach?

Follow each answer label on the arrows as drawn. Choose the end box (a rounded box with no outgoing arrow) where the path finishes.

CCA0002

WhatamIdoing, CC BY-SA 4.0, via Wikimedia Commons (File:Possible workflow for reviewers at English Wikipedia.png); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0038 Flowcharts and decision trees

A passenger is male, 8 years old and had 3 siblings or spouses aboard. Which leaf of the tree does this passenger reach?

At every question box the 'yes' branch goes down and to the left and the 'no' branch goes down and to the right, as the two labelled boxes at the top show. 'sibsp' is the number of siblings or spouses aboard. Answer with the leaf label as printed (class, then the two numbers beneath it).

CCA0032

Stephen Milborrow, CC BY-SA 3.0, via Wikimedia Commons (File:CART tree titanic survivors.png); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0039 Flowcharts and decision trees

You search for a term and an article about it already exists. You think of another term and search again; this time nothing is found and there is no related term. Which box finishes the procedure?

Follow every Yes/No label as drawn and every arrow, including the long curved arrow back to the top.

CCA0033

derivative work: Pbroks13 (talk) Wikipedia_article-creation.svg: Linuxerist, Public domain, via Wikimedia Commons (File:Wikipedia article-creation-2.svg); credit: Wikipedia_article-creation.svg, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0040 Flowcharts and decision trees

An edit request has no strong consensus, nobody has opposed it, and it proposes minor new parameters. There is some risk that the edit would be controversial, and the request has been open for one day. Which box ends the path in this chart?

Follow the labelled branches. 'Less than a few days' applies to a request open for one day.

CCA0034

Chaotic Enby, CC0, via Wikimedia Commons (File:Template edit request flowchart (part 1).svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0041 Flowcharts and decision trees

An edit request is submitted as a diff of code changes. The left-hand side of the diff does not match the existing code, but bringing the diff in sync would not break the new code. How many green action boxes does the request pass through before reaching 'Move to Part III'?

Count each green box on the path once. Blue boxes and decision diamonds are not counted.

CP_CCA0035_v1

Chaotic Enby, CC0, via Wikimedia Commons (File:Template edit request flowchart (part 2).svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0042 Flowcharts and decision trees

Inputs are A = 9 and B = 30. How many times is the box 'B <- B - A' executed before the chart reaches 'DISKWEL A'?

Translations for convenience: 'ya' = yes; 'ket' = no / otherwise; 'ket (< pe =)' = otherwise (A < B or A = B); 'ANTRE' = input; 'KAE DA 2' = go to step 2; 'DISKWEL' = display. Follow the arrows including the two return arrows that re-enter above step 2. The numbers 1-9 beside the boxes are step labels only.

CCA0036

Huñvreüs, CC BY-SA 4.0, via Wikimedia Commons (File:Euclid algorithm flowchart diagram in Breton.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0043 Flowcharts and decision trees

Start at 'Previous consensus'. Someone makes an edit, the article is then edited further, you disagree with the change, a compromise is reached and implemented, and the article is then not edited further. In total, how many times is the box 'Make an edit' visited?

Count each arrival at the box, including the first.

CCA0037

Locke Cole at English Wikipedia, CC BY-SA 3.0, via Wikimedia Commons (File:Consensus Flowchart.svg); credit: Transferred from en.wikipedia to Commons by Jayantanth using CommonsHelper., CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0044 Flowcharts and decision trees

A person taking part is not an old steward. Counting only rectangular boxes (not the ovals or the small circle), how many boxes are passed from Start to End?

Count each rectangle on the path once, including 'Steward election starts' and 'Final decisions'.

CCA0038

Created by Michel Bakni, edited and uploaded by علاء, CC BY-SA 3.0, via Wikimedia Commons (File:Flowchart of Stewards election.svg); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0045 Flowcharts and decision trees

After the initial counting box, the totals are A = 60, B = 25 and C = 15. All three candidates remain. According to the drawn procedure, what action happens next?

Follow the arrows and tests exactly as drawn, rather than using an election rule remembered from elsewhere. No candidate has withdrawn.

CCA0041

Zerodamage, CC BY-SA 3.0, via Wikimedia Commons (File:IRV counting flowchart.svg); credit: File:Alternative IRV counting flowchart.png, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0046 Causal and counterfactual reasoning

Inputs are A = 1, B = 0 and C_in = 1. Suppose the output of the upper of the two AND gates is forced to 0 whatever its inputs are. What are the circuit's outputs S and C_out?

Standard gate symbols: a D shape with a flat back is AND, a curved back with one extra curved line is XOR, a curved back is OR. A black dot marks a connection; lines that cross without a dot do not connect. Only that one AND-gate output is changed.

CCA0004

en:User:Cburnett, CC BY-SA 3.0, via Wikimedia Commons (File:Full-adder.svg); credit: Own work  This W3C-unspecified vector image was created with Inkscape ., CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0047 Causal and counterfactual reasoning

Enable E = 1 and data D = 1. Before this, the stored output Q was 0. Suppose the output of the lower AND gate is forced to 0 whatever its inputs are. What are Q and the inverted output Q-bar afterwards?

The two cross-coupled OR-with-circle gates are NOR gates. A NOR pair whose other inputs from the AND gates are both 0 keeps its previous outputs. A dot marks a connection; lines that cross without a dot do not connect.

CCA0050

Oğuz Ergin, CC BY-SA 4.0, via Wikimedia Commons (File:Flip flop D mandali tr.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0048 Causal and counterfactual reasoning

The depicted adder receives A = 0111 and B = 0001, with carry in 0. If the carry wire connecting the leftmost two boxes is forced to 0, what are s3 s2 s1 s0?

Each box is a one-bit full adder: sum is the total of its two input bits and incoming carry modulo 2; outgoing carry is 1 if at least two inputs are 1. Use the drawn bit labels and wiring to identify the affected stages.

CCA0055

Mik81, CC0, via Wikimedia Commons (File:Ripple-carry-adder.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0049 Causal and counterfactual reasoning

Inputs are p = 1, q = 1 and r = 0. The first gate (the leftmost, controlled by p) is removed so that it does nothing. What values appear on the 'carry' and 'parity' outputs?

Each vertical connection is a swap gate: a black dot on one line is the control, and the two crossed lines are swapped if and only if the control line carries 1. The two lowest input lines carry the constants 0 (upper) and 1 (lower). Gates act from left to right.

CCA0056

Bret Mulvey, CC BY-SA 4.0, via Wikimedia Commons (File:Fredkin Full Adder.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0050 Charts and quantitative comparisons

Between 2000 and 2010, which country's share of the population using the Internet increased by the largest number of percentage points?

Read the value of each country's line at 2000 and at 2010 from the vertical axis and compare the increases. Use only the plotted chart.

CCA0010: image withheld pending third-party terms review. Tested SHA256: a22ea71183cadebee83e68fbca6f345b8c2e9535198bdaeed76ba8fa2b7ec010.

Our World in Data, 'Share of the population using the Internet', data source: World Telecommunication/ICT Indicators Database - International Telecommunication Union (ITU), via World Bank (2026); CC BY, CC BY 4.0 (per OWID FAQ; third-party source terms not independently cleared). Prepared model input from the cited source; see original source for its unmodified file.

VB0051 Charts and quantitative comparisons

In 2020, about how many times larger was the United States' GDP per capita than India's?

Read both values at 2020 from the vertical axis, divide, and choose the nearest option.

CCA0059: image withheld pending third-party terms review. Tested SHA256: 29e35248e89176fbe01ecfe8a94d517cdb4340a0c95768a620b8267c2546a883.

Our World in Data, 'GDP per capita', data source: Eurostat, OECD, IMF, and World Bank (2026); CC BY, CC BY 4.0 (per OWID FAQ; third-party source terms not independently cleared). Prepared model input from the cited source; see original source for its unmodified file.

VB0052 Charts and quantitative comparisons

In 2010, how many of the other four countries used less than one third as much primary energy per person as the United States?

Read all five lines at 2010 (a labelled tick) and compare each other country with one third of the United States value.

CCA0062: image withheld pending third-party terms review. Tested SHA256: d3d6b6e05aebd49bc8b420e878e0d8437a9da92df186ec355f9e968ba4d66925.

Our World in Data, 'Primary energy use per person', data source: Energy Institute – Statistical Review of World Energy (2026) and other sources; CC BY, CC BY 4.0 (per OWID FAQ; third-party source terms not independently cleared). Prepared model input from the cited source; see original source for its unmodified file.

VB0053 Charts and quantitative comparisons

In which period did China's annual CO2 emissions first rise above those of the United States?

Find where the China line first crosses above the United States line. The horizontal axis starts in 1792; use the labelled years to estimate the crossing.

CCA0065: image withheld pending third-party terms review. Tested SHA256: e9da119c45650a1aec35833086543ed46ae7558f3c1c42f4c2175c36dc047b84.

Our World in Data, 'Annual CO₂ emissions', data source: Global Carbon Budget (2025); CC BY, CC BY 4.0 (per OWID FAQ; third-party source terms not independently cleared). Prepared model input from the cited source; see original source for its unmodified file.

VB0054 Document reconciliation

Which formula reproduces the printed TOTAL without double-counting any tax base?

Use all ten item-price lines from CARPACCIO through CREME BRULEE. GST TAXSUBTOTAL and LIQUOR TAXSUBTOTAL are separate printed lines from GST TAX and LIQUOR TAX.

CP_CCA0012_v1

Roland Tanglao from Vancouver, Canada, CC BY 2.0, via Wikimedia Commons (File:A fine bill for a fine meal!.jpg); credit: Umami Bill from July 11 2005, CC BY 2.0. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing. Re-encoded for the tested C1 input variant.

VB0055 Document reconciliation

On this supermarket receipt, treat the lines marked with an asterisk (* or *TX after the price) as the taxable items and TAX PAID as the tax charged on them. What tax rate was applied?

Use the printed prices of the starred lines only. Round to the nearest whole percent.

CCA0013

Struthious Bandersnatch, CC0, via Wikimedia Commons (File:Alexander's Supermarket receipt, late twentieth century.tif); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0056 Document reconciliation

Which item-line difference accounts for the difference between the two printed TOTAL amounts?

Compare the two tickets as shown. HT is before VAT, TVA is VAT and TTC includes VAT.

CP_CCA0111_v2

Crêperie Tante Lucie, Public domain, via Wikimedia Commons (File:Tickets de caisse de la crêperie Tante Lucie.jpg); credit: Crêperie Tante Lucie, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0057 Document reconciliation

For the printed number of litres, suppose the unit price had been 0.990 EUR per litre instead of the printed unit price. How much higher would the rounded total be than the printed total?

Decimal commas are used. Calculate the new total and round it to the nearest cent before comparing.

CP_CCA0112_v1

HIT-Tankstelle, Public domain, via Wikimedia Commons (File:Tankquittung HIT-Tankstelle Rheinbach mit Abschnitt Technische Sicherheitseinrichtung, 2020.jpg); credit: Scan by Raimond Spekking, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0058 Document reconciliation

If the customer handed over 40.00 EUR instead of the printed cash amount, how much change should they receive after the printed deduction?

Decimal commas are used. Summe is total, MwSt.-Senkung is a deduction, Bar Euro is cash and Rueckgeld is change. Keep the purchases and deduction unchanged.

CP_CCA0113_v1

Fressnapf, Public domain, via Wikimedia Commons (File:Kassenzettel Fressnapf mit ausgewiesener MwSt-Senkung und TSE Transaktionsnummer 12 2020.png); credit: Scan by Raimond Spekking, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0059 Document reconciliation

If one of the cola drinks were removed while every other purchased unit stayed the same, what would the revised total be?

Columns are product, unit price, quantity and line amount. The cola row is labelled 콜라. Prices are VAT-inclusive; subtract one printed unit price with no other repricing or adjustment.

CP_CCA0114_v1

이태리집(태백시 소재), Public domain, via Wikimedia Commons (File:Receipt taebaek restaurant IMG 2614 modified.jpg); credit: 이태리집(태백시 소재), Public domain. Prepared model input from the cited source; see original source for its unmodified file. Reviewed crop or preparation replaces the original source framing.

VB0060 Document reconciliation

If the amount before VAT increased by 10.00 CHF, keeping the VAT rate printed on this receipt, what would the new VAT amount be?

The printed total includes VAT. Compute the current tax-exclusive base from that total, add 10.00 CHF, then apply the printed rate and round VAT to the nearest cent.

CCA0115

Audrius Meskauskas (Audriusa), CC BY-SA 3.0, via Wikimedia Commons (File:ReceiptSwiss.jpg); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0061 Document reconciliation

If the same exchange direction and net rate shown on this memo were applied to MOP 525.00, how many HKD would be received?

Use the printed net rate with no extra fee. Round the result to two decimal places.

CCA0108

Solomon203, CC BY-SA 4.0, via Wikimedia Commons (File:FIRMMC exchange memo at Macau Airport 20180527.jpg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0062 Document reconciliation

What change would be due if the customer handed over 200.00 yuan instead of the printed cash amount, keeping the purchases and rounding adjustment unchanged?

合计 = total; 去零 = rounding-off deduction; 现金 = cash; 找零 = change. Subtract the printed deduction from the printed total before calculating change.

CCA0107

User:Smyth, Public domain, via Wikimedia Commons (File:Chinese Swedish rounding.jpg); credit: Own work, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0063 Document reconciliation

What is the sum of the three Total Cost amounts in this purchase-order table?

Use the printed Total Cost column. Rounding to the nearest cent.

CCA0109

Kxtar, CC BY-SA 4.0, via Wikimedia Commons (File:Invoice1633.jpg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0064 Interface state decisions

A user wants both 'Show preview on first edit' and 'Show preview before edit box' to be turned on in this preferences tab. What is the minimum change needed before pressing Save?

Clicking a checkbox toggles it. Do not change any other setting. Ignore the other tabs.

CCA0021

self, CC BY-SA 3.0, via Wikimedia Commons (File:Screenshot Preferences Βικιβιβλία 5.png); credit: www.mediawiki.org (source code), el.wikibooks.org (content), CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0065 Interface state decisions

A user wants dates written in the all-numeric style that begins with the year (like 2001-04-08 09:12:34) and does not want any time-zone change. Which action on this page achieves that?

Pressing Save stores the visible choices. Do not change fields that the goal does not mention.

CCA0020

Original uploader was Ligulem at meta Wikimedia Foundation Governance Wiki, CC BY-SA 4.0, via Wikimedia Commons (File:Date-preferences-Screenshot.png); credit: Transferred from meta.wikimedia to Commons., CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0066 Interface state decisions

A user wants to keep 'Year in Review' as it is but does not want tables in articles to be collapsed automatically. What is the smallest change to make on this screen?

A switch with its knob on the right and a blue track is on; a switch with its knob on the left is off. Tapping a switch flips it. Make no other change.

CCA0116

SChekfa-WMF, CC BY 4.0, via Wikimedia Commons (File:Android settings screen year in review.png); credit: Wikimedia Foundation Design, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0067 Interface state decisions

A mapper wants to see only the flat (non-panoramic) photos from the Panoramax overlay and keep every other overlay setting unchanged. What is the smallest change to make?

A ticked box means the item is shown. A greyed-out item cannot be changed.

CCA0118

Pajbr, CC0, via Wikimedia Commons (File:Panoramax in iD editor 20241002.png); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0068 Interface state decisions

The edit summary 'Formatting' is correct, but the user's change is not a minor edit and they still want to watch the page. What must be changed before tapping PUBLISH?

A box with a blue fill and a white check mark is ticked. Make the smallest possible change.

CCA0119

ARamadan-WMF, CC BY-SA 4.0, via Wikimedia Commons (File:Edit-summary in Wikipedia mobile app.png); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0069 Interface state decisions

The user wants to log in to de.wikipedia while keeping the program's menus in English. Apart from entering credentials, what is the smallest required change from the visible state?

Project selects the wiki site. Language selects the program's menu language. Leave unrelated settings unchanged.

CCA0125

Addshore at English Wikipedia, Public domain, via Wikimedia Commons (File:Huggle Login Form.png); credit: Transferred from en.wikipedia to Commons by IngerAlHaosului using CommonsHelper., Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0070 Image and text consistency

Which one of the following statements about this departures board is true?

The column headed 'Spår' gives the platform number. The board lists fifteen departures. Exactly one statement is true.

CCA0018

Yann Forget, CC BY-SA 4.0, via Wikimedia Commons (File:Train departure board in Göteborg railway station, Sweden .jpg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0071 Image and text consistency

Which one of the following statements about this train information display is true?

The times in brackets are the scheduled times at each calling point. The large 13:24 is the departure time shown at the top. Exactly one statement is true.

CCA0019

John Lucas, CC BY-SA 2.0, via Wikimedia Commons (File:A Borders Railway departure board - geograph.org.uk - 7744028.jpg); credit: Geograph Britain and Ireland, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0072 Image and text consistency

Which one of these statements is consistent with the scoreboard?

Exactly one statement is consistent with the picture. 'TOL' means timeouts left and 'QTR' the quarter. The team name above each score is the team that owns that score.

CCA0128

BrokenSphere, CC BY-SA 3.0, via Wikimedia Commons (File:Final score of Maryland at Cal 2009-09-05.JPG); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0073 Image and text consistency

Which one of these statements is consistent with the road sign?

Exactly one statement is consistent with the picture.

CCA0131

Leimanbhradain, CC BY-SA 4.0, via Wikimedia Commons (File:Sign-1030661, M4, Co. Kildare, Ireland.jpg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing. Resized from [1280, 960] to [1024, 768] before testing.

VB0074 Image and text consistency

Which one of these statements is consistent with the road sign?

Exactly one statement is consistent with the picture.

CCA0132

Leimanbhradain, CC BY-SA 4.0, via Wikimedia Commons (File:Sign-1040504, Celbridge, Co. Kildare, Ireland.jpg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0075 Image and text consistency

Which statement is consistent with the road sign?

The printed number beside each place is its signed distance in kilometres. Compare those displayed distances without assuming the destinations lie on one unbranched route.

CCA0133

Mucklagh, CC BY-SA 4.0, via Wikimedia Commons (File:Sign-1050251, Dingle, Co. Kerry, Ireland.jpg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0076 Image and text consistency

Which one of these statements about the chess position is consistent with the picture?

Exactly one statement is consistent with the picture. Files run a-h from left to right and ranks 1-8 from bottom to top, as the board labels show. White pieces are the light-coloured ones, black the dark ones.

CCA0130

Melv douc, CC BY-SA 4.0, via Wikimedia Commons (File:Studio de Reti.JPG); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0077 Image and text consistency

Which one of these statements is consistent with the clock's display?

Exactly one statement is consistent with the picture.

CCA0136

unknown, CC BY-SA 3.0, via Wikimedia Commons (File:Atomic clock.jpg), CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0078 Image and text consistency

Which one of these statements is consistent with the chart?

Exactly one statement is consistent with the picture. Use the plotted lines and the labelled axes; 'the end of the chart' means the right-hand end of each line.

CCA0134: image withheld pending third-party terms review. Tested SHA256: e0b7fd82c2cfdc9287e27e8f07b694c4eb8d7c9853e3e9700bf2af508fe43733.

Our World in Data, 'Child mortality rate', data source: Gapminder (2015); UN Inter-agency Group for Child Mortality Estimation (2025); CC BY, CC BY 4.0 (per OWID FAQ; third-party source terms not independently cleared). Prepared model input from the cited source; see original source for its unmodified file.

VB0079 Image and text consistency

Which one of these statements is consistent with the chart?

Exactly one statement is consistent with the picture. Use the plotted lines and the labelled axes.

CCA0135: image withheld pending third-party terms review. Tested SHA256: 794c70c12da3475b712a66261c30472fb50d78323899164bd98f2be8d51d09c2.

Our World in Data, 'Maternal mortality ratio', data source: UN MMEIG (2023) and other sources; CC BY, CC BY 4.0 (per OWID FAQ; third-party source terms not independently cleared). Prepared model input from the cited source; see original source for its unmodified file.

VB0080 Attribute binding and set logic

To the right of the tall vertical wooden post, small ancient glass inlays are mounted in three general vertical columns (left, middle, and right). Which motif is depicted on the inlay fragment that is both located in the rightmost column and has a rounded blue body?

Bind both properties (rightmost column position and rounded blue body) to the same single fragment before identifying its depicted motif.

GE_A0001

Gary Lee Todd, Ph.D. / Wikimedia Commons / CC0 1.0, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0081 Attribute binding and set logic

In the main row of upright pens, some pens are uncapped to expose a pointed metal fountain-pen nib at the top, while others remain capped. Which barrel description belongs to an upright pen that both has an exposed classic metal fountain-pen nib at the top and stands to the left of the black-and-silver 'MORISON' pen?

Consider only the upright pens standing side-by-side in the main vertical row. An exposed classic fountain-pen nib is an open split metal nib protruding from a dark grip section at the top end of the pen.

GE_A0002

Billjones94 (Yogabrata Chakraborty) / Wikimedia Commons / CC BY-SA 4.0, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0082 Conditional counting

How many billiard balls are visibly resting completely outside the triangular rack in this photograph?

Count only balls visible within the photograph and completely outside the rack's interior perimeter. Do not infer objects beyond the image boundaries.

GE_A0003

File:Billiard Balls and Rack.jpg / Wikimedia Commons (creator verification pending), CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0083 Conditional counting

Several dozen coins are scattered on the dark ribbed mat in front of the cat. How many of the visible coins are bimetallic—specifically showing a silver-colored inner core surrounded by a yellow-gold outer ring?

Count only coins on the mat that visibly display both a silver-colored central disc and a contrasting yellow-gold outer ring. Exclude every coin cut off by an image border, as well as solid yellow-gold or copper-colored coins. Partial overlap by another object is allowed only if both qualifying materials remain clearly visible.

GE_A0004

Continentaleurope / Wikimedia Commons / CC BY-SA 4.0, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0084 Conditional counting

This diagram shows a small trefoil knot at the upper left, a larger rectangular-loop knot diagram at the lower left, and a deformed version of that larger diagram on the right after the red-arrow move. How many strand crossings appear in the small upper-left knot, and how many strand crossings appear in the right-hand diagram?

A strand crossing is a point where one white-filled black-bordered ribbon passes directly over another, indicated by a break in the under-passing ribbon's black border lines.

GE_A0023

AnonMoos / Wikimedia Commons / Public Domain, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0085 Spatial relationships

Which square is in the same rank as the white rook and the same file as the white knight?

Use the file letters and rank numbers printed on the board. A rank is a horizontal row and a file is a vertical column.

GE_A0005

File:ArabicMate.svg / Wikimedia Commons (creator verification pending), CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0086 Rotation and reflection

Suppose the left four-sphere figure is reflected horizontally across a vertical line passing through the central black point (a left-to-right mirror flip in the image plane, with no vertical flip). Compared with the actual right-hand figure shown in the diagram, which pair of vertex sphere colors would remain swapped?

A horizontal reflection across a vertical line reverses left-right positions while keeping top-bottom positions unchanged in the two-dimensional image.

GE_A0008

File:3D point reflection.png / Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0087 Geometry, folding and fit

When this net is folded into a closed cube, which square is opposite the central square that has another square attached on both its left and its right?

Fold along the shared square edges without cutting, stretching or overlapping faces. Identify positions in the pictured flat net.

GE_A0009

File:Hexahedron flat.svg / Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0088 Inspection against a reference

Image 1 is a reference photograph of four candles. In Image 2, which candle has a clearly visible cluster of dark particles on its wax surface above the wick, unlike the reference candles?

Identify a visible surface difference only; do not infer product safety or hidden damage. Positions refer to the two-by-two arrangement in Image 2.

GE_A0013

Zou et al., Spot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentation (VisA Dataset), Amazon Science, CC BY 4.0, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file.

GE_A0014

Zou et al., VisA Dataset, Amazon Science, CC BY 4.0, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0089 Matching across images

Image 1 shows a frontal view of an ancient painted amphora, and Image 2 shows the same amphora viewed from an angle centered on one of its two vertical handles. By matching the painted neck scene, vertical geometric border strips, and lower body figures visible next to the handle in Image 2 against Image 1, which side of Image 1 corresponds to the handle centered in Image 2?

Match specific visible painted features—such as the direction the neck figure faces, the vertical inscription beside its back, and the vertical border strips—between the two views.

GE_A0017_V1

George E. Koronaios / Wikimedia Commons / CC BY-SA 4.0, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

GE_A0017_V2

George E. Koronaios / Wikimedia Commons / CC BY-SA 4.0, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0090 Insufficient evidence

The lower line of the white roadside nameplate is partly covered by vegetation. Which of these complete names can be uniquely established from the visible lettering?

Use only the visible image. Do not use the source title, a map, location knowledge or an inferred completion of covered letters.

GE_A0021

Kenneth Allen / Geograph Britain and Ireland / Wikimedia Commons / CC BY-SA 2.0, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0091 Insufficient evidence

Which royal cypher can be established from the visible features of the red postbox in this photograph?

Use only a readable emblem on the postbox itself. Do not infer a monarch from the location, casing style, source caption or surrounding buildings.

GE_A0022

Geograph 1623572 / Wikimedia Commons / CC BY-SA 2.0 (author verification pending), CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0092 Perspective and knowledge tracking

In this polar visibility chart around a haul truck (where 0 degrees is straight ahead, angles increase clockwise around the outer circle, dashed circles mark distances from 10 ft to 50 ft, grey regions show where the operator cannot see a 6-foot-tall person, and white regions show where the person is visible), at a distance of 40 feet from the center, at which pair of bearings can the operator see a 6-foot-tall person?

Read the bearing angles around the 50-foot outer circle and locate the intersection with the 40-foot dashed concentric circle. Both bearings in the chosen pair must lie in a white (visible) region.

GE_A0020

CDC / NIOSH / Public Domain, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0093 Attribute binding and set logic

Which object is below the haystack and to the left of the large upright spoked wheel?

Use image directions. Both position requirements must describe the same object.

NC_A0001

Sebastian Ballard; File:Assortment of farm implements - geograph.org.uk - 1012935.jpg; CC BY-SA 2.0; via Wikimedia Commons, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0094 Conditional counting

Compare the blue circles with all the red and orange circles combined. Which group is larger, and by how many circles?

Count both shades of red in the red group. Count each circle once; the printed 99 is not a color-group count.

NC_A0007

Lluket24; File:Corts 2011.jpg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0095 Spatial relationships

What is the minimum number of unit grid steps from the white king to the black rook farthest to the image left?

A step moves one square horizontally or vertically. Ignore chess movement rules, other pieces and obstacles. Use the printed coordinates, including negative numbers.

NC_A0008

LaundryPizza03; File:Infinite chess.png; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0096 Rotation and reflection

Imagine rotating the whole photograph 90 degrees clockwise. Relative to the cream cube marked K near the upper-right, where would the red cube with the large 8 be after that rotation?

Use positions on the rotated image. Rotate both dice together; do not turn either die separately.

NC_A0004

Ironmaidenfun; File:Assorted dice for games, 2024.jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0097 Inspection against a reference

Compared with the soft shadows around the reference capsules in Image 1, which capsule in Image 2 has an irregular, sharply bounded black patch extending around its lower edge?

Identify the visible pattern only. Do not infer leakage, chemical contents or product safety.

NV_A0004

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0005

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0098 Inspection against a reference

After matching the printed T and R labels between the two photographs, which visible part is missing in Image 2 compared with Image 1?

The board orientations differ. Match the labels, not just the image-left and image-right positions. Do not infer electrical function.

NV_A0010

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0012

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0099 Inspection against a reference

Which region of the object in Image 2 has a rough pit that is not present on the smooth reference object in Image 1?

Judge the visible surface, not a proposed cause or whether the object is safe to use.

NV_A0018

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0020

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0100 Inspection against a reference

For Image 2, which pair of observations is supported when compared with the reference wheel in Image 1?

Check these separately: whether the outside rim has a break, and whether an additional fragment crosses the openings inside the wheel. Ignore ordinary rotation.

NV_A0021

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0022

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0101 Inspection against a reference

Which pasta piece in Image 2 has a clear pale fracture line crossing its curved body, unlike the unfractured reference pieces in Image 1?

Positions refer to the two-by-two arrangement in Image 2. A change in orientation alone is not a fracture.

NV_A0027

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0028

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0102 Inspection against a reference

Which component has changed orientation relative to the circuit board in Image 2?

Match the blue square adjuster and the three-pin header first. Compare component directions relative to the board, not the camera. Do not infer electrical performance.

NV_A0030

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0031

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0103 Inspection against a reference

Which pair of parts is present in the reference board in Image 1 but absent in Image 2?

Align the boards using the USB connector and the IN+/IN- labels. Count visible components, not their empty solder pads.

NV_A0033

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0035

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0104 Inspection against a reference

Which wall difference is visible in Image 2 compared with the reference tube in Image 1?

Inspect the visible sidewall only. Do not infer hidden damage or how the opening was made.

NV_A0036

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0016

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0105 Sequential state tracking

Across Images 1, 2 and 3 in that order, how does the visible illuminated fraction of the solar disk change?

Compare the fraction of the disk that is lit, not color, exposure or the amount of black margin around each crop. These are ordered crops from one existing photographic sequence.

NS_A0001_F1

Douglas Duncan/IAU OAE (CC BY 4.0); File:Smartphone Partial Eclipse Combined (ann23043r).jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NS_A0001_F2

Douglas Duncan/IAU OAE (CC BY 4.0); File:Smartphone Partial Eclipse Combined (ann23043r).jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NS_A0001_F3

Douglas Duncan/IAU OAE (CC BY 4.0); File:Smartphone Partial Eclipse Combined (ann23043r).jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0106 Geometry, folding and fit

When this net is folded into its closed convex solid, how many vertices does the solid have?

Every drawn polygon becomes one face. Every completed edge is shared by exactly two faces. You may use V - E + F = 2. Count the actual panels and their sides; no panel disappears or splits.

NC_A0009

Timeroot; File:Elongated Pentagonal Pyramid Net.svg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0107 Geometry, folding and fit

After folding the pictured net, how many yellow triangular faces meet at each vertex on a grey base?

The grey polygons form disjoint parallel bases. The equal yellow triangles close into the repeating band between them, with no additional vertices away from the bases.

NC_A0010

Timeroot; File:Generalized antiprisim net.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0108 Geometry, folding and fit

How many pairs of cut boundary edges must be joined to close this one-piece net?

Every black shared edge between two panels is already a hinge. Each remaining outside boundary edge must be joined to exactly one other boundary edge. Do not count a hinge as a new join.

NC_A0011

A2569875; File:Net of self-dual form tetrahedrally diminished dodecahedron.svg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0109 Rotation and reflection

Imagine rotating the whole net 180 degrees in the image plane. Relative to the red pentagon that is currently bottommost, where would the red pentagon that is currently topmost lie after the rotation?

Track the same two faces through the rotation. Use directions on the resulting image, not directions on a folded solid.

NC_A0012

Cmglee; File:Polyhedron truncated 20 net compact.svg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0110 Constraint satisfaction

Starting at the king symbol, which route satisfies all three conditions: exactly four moves, no square visited twice, and a sum of 11 across the numbered squares entered?

Each named direction moves one square, including diagonals. Up is image-up. The starting square counts as visited but contributes zero. Remain inside the drawn grid.

NC_A0006

w:de:Benutzer:Gunther; File:King walks.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0111 Perspective and knowledge tracking

Initially every watcher knows that all three sites display red flags. The rightmost watcher then changes that site to green. A watcher at each tower base observes continuously, with no messages. Which watchers can know that the rightmost flag is now green?

A straight segment between two tower bases is visible only if it does not cross grey terrain. Watchers learn later changes only by seeing or performing them and remember earlier observations. Ignore the green radio-link lines.

NC_A0015

Micogen; File:Line of sight mw transmission.svg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0112 Inspection against a reference

Which pair of pieces in Image 2 both has conspicuous dark purple or black spots on the broad curved surfaces, unlike the reference pieces in Image 1?

Compare visible surface marks. Do not count background texture, shadows or the dark hollow cut ends. Positions refer to Image 2.

NV_A0024

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NV_A0038

VisA dataset, Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang and Onkar Dabeer, 2022, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0113 Sequential state tracking

Which visible change occurs between Images 2 and 3 in this ordered frog-jump sequence?

Track the main frog. Describe the pictured limb and body positions only; do not predict forces or infer an unseen frame.

NS_A0002_F1

Li et al;; File:Frame-Sequence-of-Rana-dybowskii-Jump.jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NS_A0002_F2

Li et al;; File:Frame-Sequence-of-Rana-dybowskii-Jump.jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NS_A0002_F3

Li et al;; File:Frame-Sequence-of-Rana-dybowskii-Jump.jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0114 Sequential state tracking

Which sequence describes the main frog's visible hind-leg configurations across Images 1, 2 and 3?

Use the supplied image order and the coarse pictured configurations, not exact joint angles or inferred contact with the floor.

NS_A0003_F1

Li et al;; File:Frame-Sequence-of-Xenopus-Laevis-Jump.jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NS_A0003_F2

Li et al;; File:Frame-Sequence-of-Xenopus-Laevis-Jump.jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NS_A0003_F3

Li et al;; File:Frame-Sequence-of-Xenopus-Laevis-Jump.jpg; CC BY 4.0; via Wikimedia Commons, CC BY 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0115 Rotation and reflection

Keeping the narrow projecting neck at the top, which transformation best matches the outline of the left seal photograph to the outline of the right one?

Allow translation and small scale differences. Compare the asymmetric outer silhouette, not the surface cracks or the ruler. Do not turn the neck downward.

NC_A0021

Birmingham Museums Trust, Caroline Johnson, 2007-07-05 13:48:13; File:Medieval - Post Medieval bag seal (front and back views) (FindID 182369).jpg; CC BY-SA 2.0; via Wikimedia Commons, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0116 Matching across images

Which region in the right drawing corresponds to the broad layered hat crown at the top of the left drawing?

Match the concentric curved bands and the irregular straw fringe. Do not match by the color of the long hanging ribbons alone.

NC_A0020

Auguste Félix; File:Two Views (Side and Front) of a Hat on a Stand (IA mma two views side and front of a hat on a stand 387445).jpg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0117 Conditional counting

How many of the bird's feet are visibly in contact with the branch?

Count feet, not individual toes. Use visible contact only; do not infer flight, motion or hidden legs.

NC_A0025

Francis C. Franklin; File:Blue tit front and back-lit.jpg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0118 Geometry, folding and fit

After this net is folded into its closed solid, what do the red and yellow-green triangular faces share?

Fold along the polygon edges without cutting, stretching or overlapping panels. The two colored triangles remain distinct faces.

NC_A0033

IcosaMarty343; File:Triangular prism net.png; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0119 Geometry, folding and fit

After folding, how many yellow faces share an edge with both green faces?

Close the pictured prism without overlapping panels. Blue outlines define face edges; ignore faint white rendering hairlines inside a green panel.

NC_A0034

A2569875; File:Net of pentagonal prism.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0120 Geometry, folding and fit

When this net is folded, how does the total yellow surface area compare with the total pink surface area?

All pictured triangular faces are congruent. No face overlaps another in the completed solid.

NC_A0035

David Eppstein; File:Triaugmented triangular prism (symmetric net).svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0121 Geometry, folding and fit

Using other pieces from this dissection, which choice can exactly cover the central purple square?

Pieces may be moved, rotated and reflected, but not cut, stretched or overlapped. Match both area and shape, not area alone.

NC_A0037

Sarang; File:Tangram diagram.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0122 Perspective and knowledge tracking

The watcher at the white point initially records both flags as red. Flag A at(10% width,15% height) changes to green; flag B at(50% width,25% height) changes to blue. With no messages, what colors remain in the watcher's record?

Coordinates start at the image top-left. Yellow is visible to the white point; grey is hidden by the black barriers. A recorded color changes only when the watcher sees that flag change.

NC_A0029

Dllu; File:Visibility polygon for a point among segments.svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0123 Perspective and knowledge tracking

At moment1 a marker at(50% width,20% height) is red. At moment2 it is at(75%,20%) and turns blue. At moment3 it is at(50%,80%) and remains blue. When can the watcher at the white point first record it as blue?

Coordinates start at the top-left. The watcher observes only at the three stated moments, with no observations between them. Yellow is visible and grey hidden. Only a visible marker can update its recorded color.

NC_A0030

Dllu; File:Visibility polygon for a point in a simple polygon 1.svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0124 Matching across images

Which side wall of the building-like case in Image1 is revealed in Image2, in addition to its front?

Define the front by the large central lower door and curved balcony in Image1. Left and right refer to that front-on view. Match those features before deciding.

NM_A0001_V1

National Museum of American History; perfume case, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

NM_A0001_V2

National Museum of American History; perfume case, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0125 Matching across images

Which button in the horizontal row of six in Image2 matches the detailed button in Image1?

Count positions from image-left. Match the arrangement of carved fruit, flower and leaves, not just the ivory color or the rim.

NM_A0002_V1

Cooper Hewitt, Smithsonian Design Museum; Button, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

NM_A0002_V2

Cooper Hewitt, Smithsonian Design Museum; Button, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0126 Matching across images

Which of the four figures in Image2 belongs to the object shown alone in Image1?

Order the figures by their head positions from image-left to image-right. Match the raised arm, cup position and surrounding scrollwork; do not rely on ceramic color alone.

NM_A0003_V1

Cooper Hewitt, Smithsonian Design Museum; Taperstick (One of Four), CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NM_A0003_V3

Cooper Hewitt, Smithsonian Design Museum; Taperstick (One of Four), CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0127 Attribute binding and set logic

What color is the large oval painted mark that is above the protruding face and between the upper red roundel and the handle?

Use image positions and bind all location requirements to the same painted mark. Do not count the many small leaf-shaped marks as large medallions.

NM_A0004_V1

Cooper Hewitt, Smithsonian Design Museum; Jug, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0128 Insufficient evidence

Which geographic compass direction does the flower's stem extend toward in this photograph?

Geographic north/east/south/west are not image-up/right/down/left. Use only the supplied picture, not a source location, metadata or an assumed camera heading.

NC_A0024

Sharon Mollerus; File:Rose in Three Views (Close) (2536313844).jpg; CC BY 2.0; via Wikimedia Commons, CC BY 2.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0129 Conditional counting

How many elongated brown basket-weave pendants with gold tips are visible across the pair of earrings?

Exclude the nearly round brown connector directly under each pair of leaves, the metal leaf mounts, and the hooks. Count each elongated pendant once.

NM_A0008_V1

Cooper Hewitt, Smithsonian Design Museum; Pair of earrings, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0130 Matching across images

Which part in Image2 corresponds to the small curl projecting to the left of the coil at mid-height in Image1?

Match the physical attachment to the coil and distinguish the short curl from the long outer arm and the feet. Camera direction changes between photographs.

NM_A0012_V1

Cooper Hewitt, Smithsonian Design Museum; Candle holder, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NM_A0012_V2

Cooper Hewitt, Smithsonian Design Museum; Candle holder, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0131 Matching across images

In Image2, which image direction leads from the two threaded posts toward the dial face shown front-on in Image1?

Match the circular dial rim to the rim in the side view. Use image directions, not a compass heading. No judgment about instrument operation is required.

NM_A0014_V2

National Air and Space Museum; Indicator, Cylinder Head Temperature, Japanese, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NM_A0014_V3

National Air and Space Museum; Indicator, Cylinder Head Temperature, Japanese, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0132 Matching across images

Which figure in Image2 matches the figure in Image1 in both clothing and pose?

Match visible garment shapes, leg positions and arm positions. Similar colors alone do not establish a match; a no-match answer is allowed.

ND_A0001

Cooper Hewitt, Smithsonian Design Museum; Pleated fan, CC0. Prepared model input from the cited source; see original source for its unmodified file.

NM_A0006_V2

Cooper Hewitt, Smithsonian Design Museum; Pleated fan, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0133 Matching across images

Which description correctly matches features between these two spoon photographs?

Compare shapes and decoration separately from apparent brightness. A dark reflection is not by itself evidence that a feature disappeared.

NM_A0009_V1

Cooper Hewitt, Smithsonian Design Museum; Tea Caddy Spoon, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

NM_A0009_V2

Cooper Hewitt, Smithsonian Design Museum; Tea Caddy Spoon, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0134 Conditional counting

How many large tassels hang from the two outermost vertical cords combined?

Use the single far-left cord and single far-right cord. Exclude tassels on inner cords, the short fringe along the upper valance, and small architectural ornaments.

NM_A0013_V1

Cooper Hewitt, Smithsonian Design Museum; Window shade, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0135 Insufficient evidence

How many painted flower groupings are there in total on the saucer's inner surface?

Count groupings on the saucer, not on the cup. Use this photograph alone. Do not assume that hidden decoration repeats visible decoration or that the cup is transparent.

NM_A0010_V1

Cooper Hewitt, Smithsonian Design Museum; Cup and Saucer, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0136 Rotation and reflection

Imagine rotating this entire image 90 degrees counterclockwise. In which direction from the shield-shaped head would the long handle then extend?

Rotate the image in its plane, not the object in depth. Use directions on the rotated image.

NM_A0011_V1

Cooper Hewitt, Smithsonian Design Museum; Handscreen, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0137 Attribute binding and set logic

Which motif is above the green bird on the right of the verse and to the left of the pale blue hanging bell-shaped flowers?

Bind both position requirements to the same complete motif. Use image directions and ignore small border ornaments.

NM_A0016_V2

Cooper Hewitt, Smithsonian Design Museum; Sampler, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0138 Insufficient evidence

Which route from node4 to node1 is guaranteed to have the least total cost?

Every drawn connection is undirected and has a fixed but unreported strictly positive cost. Route cost is the sum of its edge costs; node labels are names, not costs. Do not assume equal costs.

CCA0030

Tavin, Public domain, via Wikimedia Commons (File:6n-graf-2-clique.svg); credit: File:6n-graf-clique.svg, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0139 Insufficient evidence

The observed outputs are S=1 and Cout=0. What does that establish about input A?

Inputs are binary. A gate with the extra curved input line is XOR; the other gates are AND. A black dot connects crossing wires; an unmarked crossing does not. No gate is faulty and all shown connections apply.

CCA0057

inductiveload, Public domain, via Wikimedia Commons (File:Full Adder XOR.svg); credit: Own drawing, Inkscape 0.43, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0140 Insufficient evidence

Starting at q0, three inputs produce the output sequence000. Which state must the machine be in after the third input?

An edge label input/output emits the shown output while taking that transition. The input sequence contains only a and b and is not given. Follow arrowheads.

CCA0070

Aritzh, CC BY-SA 4.0, via Wikimedia Commons (File:MealyMachine.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0141 Perspective and knowledge tracking

Alice sees the first decision answered Yes, then sees nothing more. Bob sees only the final Replace bulb box reached. Who can know that the lamp was plugged in?

Both know this chart is followed exactly once from its start, without changing any answer during the traversal. They exchange no messages and have no other observations. Knowledge means true in every traversal compatible with that observer's evidence.

CCA0040

svg by Booyabazookaoriginal png by Wapcaplet, CC BY-SA 3.0, via Wikimedia Commons (File:LampFlowchart.svg); credit: vector version of Image:LampFlowchart.png, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0142 Sequential state tracking

From panel4 through panel5 to panel6, which movement sequence does the grey tool make relative to the two pads?

Track the grey tool, not the yellow wire or panel positions on the page. Compare its location relative to the left and right pads in each stage.

NC_A0028

Retired electrician; File:Ball-wedge sequence.svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0143 Geometry, folding and fit

After this net is folded into its closed prism, what is the fewest number of colors needed so that faces sharing an edge always have different colors?

You may recolor every face; the current yellow and green are only panel identifiers. Faces meeting only at a vertex may share a color. Count the actual lateral faces and how they join when folded.

NC_A0039

A2569875; File:Net of heptagonal prism.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0144 Rotation and reflection

To make a left-right mirror arrangement of this rabbit, which colored shape has a mirrored outline that cannot be obtained by rotating that piece in the table plane?

Treat the cutouts as ideal standard tangram pieces and ignore hand-cut roughness, shading and crayon strokes. Pieces may move and rotate but cannot be turned over. Compare each piece with the mirror of its own outline.

NC_A0038

Cecilia2611; File:Coniglio Tangram.jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0145 Rotation and reflection

Can the entire left cube-chain shape be made identical to the entire right one by a rigid rotation in three-dimensional space?

Read the cubes as equal units and the straight runs as perpendicular. Preserve the connected chain, including its unique four-cube end. Rotation is not reflection and pieces cannot be detached.

NR_A0001

File:Mental rotation task (diagram) (cropped).jpg / Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0146 Insufficient evidence

Which of the location-requesting apps have actually been granted location permission?

Standort means location. Use the displayed permission summary alone; do not assume app names, alphabetical order, installation order or which apps normally need location.

CCA0122

Futurwiki, Public domain, via Wikimedia Commons (File:App berechtigung.png); credit: Screenshot, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0147 Perspective and knowledge tracking

The interval is4. Alice and Bob record the start and three completed decrement steps. A kick occurs after decrement2 and before Sleep3, but only Alice sees it. What counter values do their records contain after decrement3?

Each observer updates a record using only operations they see, with the effects defined by the chart. Both see all three decrements, Alice also sees the kick, and there are no other operations. Ask for recorded values, not whether Bob's record is currently true.

CCA0044

Lambtron, CC BY-SA 4.0, via Wikimedia Commons (File:Software watchdog timer flowchart.svg); credit: Own work, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0148 Perspective and knowledge tracking

Who can establish that the shortest-route cost from d to b is at most20: Alice, who reads only d-c and c-b, or Bob, who reads only d-c, c-a and a-b?

Both know the full undirected topology and that all costs are positive. Each knows only the numeric edge costs they personally read from this image; unread costs remain unknown to that observer. They exchange no messages. A known route of cost20 or less is enough to establish the upper bound.

CCA0089

Daniel Mietchen, CC0, via Wikimedia Commons (File:Simple graph example for illustration of Bellman-Ford algorithm.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0149 Perspective and knowledge tracking

A run starts at S0 and takes exactly three transitions. Alice sees only that the first input was10. Bob sees only that the third output was0. Who can know the final state is S5?

Labels are input/output. Each observer knows the whole diagram and the run length, but only the stated observation from this run. They do not exchange messages. Knowledge must hold for every allowed three-transition run consistent with that observer's evidence.

CCA0077

Pluke, CC0, via Wikimedia Commons (File:CPT-FSM-Mealy-02.svg); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0150 Attribute binding and set logic

Which fruit is above the large round fruit near the center and horizontally between the two prominent pear-shaped fruits?

Use image positions. The left pear is high on the left and the other pear is on the right; bind both relations to one object.

NC_A0041

AnonymousUnknown author; File:Bowl of Fruit.png; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0151 Conditional counting

How many orange-colored round fruits are visible in the bowl?

Exclude green and brown-green fruits even where partly hidden. Count each visible orange-colored fruit once, not each bright patch on a fruit.

NC_A0043

PaulCharlesSmith; File:Bowl of fruit.jpg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0152 Rotation and reflection

Imagine rotating this whole photograph90 degrees counterclockwise. Relative to the center of the brown onion pile, where would the Spanish Onions label then be?

Rotate the image in its plane without reflecting it. Track the same pile and front label. Use directions on the rotated photograph.

NC_A0044

Andy Li; File:Spanish Onions - The Fruit Bowl, Brighton 2024-03-21.jpg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0153 Spatial relationships

From the center of the orange carrot toward the center of the large dark-green courgette, which pair of robot-relative directions applies?

For this task the robot's forward direction is image-right, and its right direction is image-down. Use the centers of the visible objects, not their long-axis directions.

NC_A0045

Salicyna; File:Allium cepa Karmen and other vegetables 2023-07-20 8383.jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0154 Spatial relationships

At the crossing near the image center-right, what is the visible depth order of the basket handle and the pale vegetable?

Judge the local overlap where the upright handle crosses the large pale vegetable. Do not infer the order of objects elsewhere in the basket.

NC_A0046

George Chernilevsky; File:Basket with vegetables 2017 G1.jpg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0155 Attribute binding and set logic

Which tool lies below the circuit board and above the notebook, rather than being held in the person's hand?

Bind the two position conditions and the not-held condition to the same object. Use image directions; no claim about whether this is a sensible repair method is required.

NC_A0047

Shixart1985; File:Person working on a computer circuit board with tools at a desk.jpg; CC BY 2.0; via Wikimedia Commons, CC BY 2.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0156 Attribute binding and set logic

Which mounted artifact lies directly below the left leaf-shaped arrowhead and directly to the right of the large pale circular scraper?

The leaf-shaped arrowheads are the two narrow pieces at the upper-right. Use the left one of those two, not either forked arrowhead at the upper-left.

NC_A0048

The Portable Antiquities Scheme, Chris Edbury, 2012-06-15 15:29:25; File:An old collection of arrowheads, knives and scrapers mounted on card (FindID 507979).jpg; CC BY-SA 2.0; via Wikimedia Commons, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0157 Sequential state tracking

Scanning the displayed stages from left to right, which transition first introduces a visible white fluffy center?

Compare the pictured stages in their given order, not a remembered botanical life cycle. A pale green fuzzy outside on an unopened bud is not a white fluffy center.

NC_A0049

DrunkBicyclist; File:Creosote Bush Flower Stages.jpg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0158 Sequential state tracking

Reading just the six hanging-case pictures left-to-right across the upper small row, then the lower small row, which is the first whose broad lower section is mostly dark rather than pale and striped?

Ignore the large caterpillar picture and the final spread-wing adult. Compare the broad lower section of each hanging case, not the orange spots on its narrow upper body.

NC_A0050

Nick Hobgood; File:Acraea zetes caterpillar to pupae to butterfly metamorphosis by Nick Hobgood.jpg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0159 Geometry, folding and fit

At stage3, one outer paper flap is opened to form the cone in stage4. How many paper layers are on the two sides of that cone?

Start with the single paper sheet drawn at stage1. Follow the shown folds, each placing one half over the other. Open one outer flap, not a separation through the middle of the layer stack. Do not tear or add paper.

NC_A0051

Percy Cyril Lesley Thorne; File:Paper filter folding and filtration.JPG; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0160 Insufficient evidence

Which of the two square faces is closer to the observer in the depicted three-dimensional wireframe cube?

The two squares are the lower-left square and the upper-right square linked by diagonal edges. Use depicted depth cues, not a habitual preferred interpretation of a familiar drawing.

NC_A0052

BenFrantzDale; File:Necker cube.svg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0161 Perspective and knowledge tracking

A pilot first sees a red beacon at B=(95% width,84% height). It moves to A=(75%,78%) and turns green there. A shore observer sees the whole event. With the beacon still at A, who can know its light is green?

Coordinates start at the top-left. For this task the red water area is hidden from the pilot and other water is visible. The pilot cannot see through the ship, receives no messages and does not know a color change occurred while hidden. Both observers initially know the light was red.

NC_A0054

Blind_spot_(watercraft).JPG: KVDP derivative work: Blleininger (talk); File:Blind spot (watercraft-2).svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0162 Perspective and knowledge tracking

Mira initially sees a white spot at the center of the glossy black face shown on the left. The fragment is then turned to the right-hand photograph's orientation. The spot turns blue while it stays in that orientation. What remains in Mira's color record?

The spot is a hypothetical mark at the center of that face. The fragment is opaque; Mira sees only the surface presented in the photograph and cannot look around its edges. She receives no message and updates her record only when she sees the spot.

NC_A0022

Birmingham Museums Trust, Caroline Johnson, 2006-05-08 14:03:17; File:Post Medieval vessel sherd (front and back) (FindID 131739).jpg; CC BY-SA 2.0; via Wikimedia Commons, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0163 Attribute binding and set logic

Which bathtub both has an obvious break in its near rim and contains greenish liquid?

Bind both visible conditions to the same bathtub. Positions are ordered by distance from the camera; do not infer whether a tub is usable or safe.

NC_A0003

Mick Garratt; File:Assortment of Baths, Redpoint Farm - geograph.org.uk - 531611.jpg; CC BY-SA 2.0; via Wikimedia Commons, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0164 Attribute binding and set logic

Which label identifies the two long folded pieces directly beside the pair of narrow flat center strips in the lower assembly layout?

Use all three properties: long, folded, and immediately outside the central flat strips. Read the labels from the pictured layout, not from clothing terminology.

NC_A0059

HLHJ; File:Kosode assembly, exploded view.svg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0165 Conditional counting

How many rounded reddish-brown end ornaments are visible on this purse in total?

Count only the ball-like ornaments attached to green leafy extensions at the two outer ends. Exclude tiny flower centers across the mesh and the two sliding rings at the middle.

NM_A0007_V1

Cooper Hewitt, Smithsonian Design Museum; Miser's purse, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0166 Conditional counting

How many pips are on the large front-facing faces of the two purple cubes with yellow pips, in total?

Ignore printed numerals, other-colored dice and the white spherical die. Each face uses a standard1-to6 die pip pattern. Include a partly hidden corner pip if the remaining pattern forces it, but do not invent a hidden center pip.

NDICE_A0001

Benny Mazur; File:Assorted dice for games at GenCon 2009.jpg; CC BY 2.0; via Wikimedia Commons, CC BY 2.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0167 Spatial relationships

Can the two rings be completely separated by moving and rotating them in three-dimensional space without opening either ring?

Both rings are physically closed and cannot pass through solid material. White gaps at the crossings indicate which ring is hidden behind the other, not actual cuts. Read both crossings before deciding.

NC_A0060

Lorc; File:Linked-rings - Lorc - game-icons.svg; CC BY 3.0; via Wikimedia Commons, CC BY 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0168 Spatial relationships

In the lower illustration, how are the large foreground boat and the smaller distant boat viewed?

End-on means looking approximately along a boat's length; broadside means seeing its length from the side. Judge hull shape and symmetry, not apparent size alone. The two illustrations need not show the same individual boat.

NC_A0040

La Perouse, Jean-Francois de Galaup, comte de, 1741-1788 Blondela Heath; File:(Two views). Japanese Boat(s). Published as the Act directs Novr. 1st 1798, by G.G. & J. Robinson, Pater Noster Row (IA dr two-views-japanese-boats-published-as-the-act-directs-novr-1st-1798-0414059).jpg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0169 Sequential state tracking

A visual gate triggers after seven consecutive pictured ball positions whose centers are in the upper half of the photograph. Which bounce arc triggers the gate?

Test the taller left arc and shorter right arc separately, in left-to-right order. Use ball centers, not their upper edges, and the horizontal middle of the whole photograph. The overlapping positions near a peak still count separately. Do not infer unpictured positions.

NC_A0055

MichaelMaggs Edit by Richard Bartz; File:Bouncing ball strobe edit.jpg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0170 Sequential state tracking

At the highest pictured position in the moving diver's sequence, which body configuration is shown before the later head-down entry?

Track the diver in striped blue-and-white trunks, not the stationary people. Compare the actual pose at the greatest height; different diving styles could plausibly use the other configurations, so do not choose from a generic remembered dive sequence.

NC_A0057

Chris Hunkeler from Carlsbad, California, USA; File:Diving Sequence (41497552975).jpg; CC BY-SA 2.0; via Wikimedia Commons, CC BY-SA 2.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0171 Sequential state tracking

A running counter starts at0. Process Images1,2 and3 in order: add1 when the broad flukes are near-black and mottled; reset the counter to0 when they are mostly uniform grey. What is the final counter value?

Classify the broad fluke surfaces, not water, spray or the narrow stalk. Apply the update separately at every frame. The counter describes a visual monitoring rule, not a biological change in skin color.

NW_A0001_F1

Jules Verne Times Two; File:Sequence of sperm whale (Physeter macrocephalus) diving, São Miguel Island, Azores, Portugal (PPL1-Corrected).jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

NW_A0001_F2

Jules Verne Times Two; File:Sequence of sperm whale (Physeter macrocephalus) diving, São Miguel Island, Azores, Portugal (PPL1-Corrected).jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

NW_A0001_F3

Jules Verne Times Two; File:Sequence of sperm whale (Physeter macrocephalus) diving, São Miguel Island, Azores, Portugal (PPL1-Corrected).jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0172 Matching across images

The thin branch ending near the left interior rim in Image2 corresponds to which branch end in Image1?

Use the blue bow and the large flower clusters to align the decorative pattern across the different camera views. Divide Image1 around the center of the dish; do not assume image-left stays object-left.

NM_A0015_V2

Cooper Hewitt, Smithsonian Design Museum; Dish (One of Six), CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

NM_A0015_V3

Cooper Hewitt, Smithsonian Design Museum; Dish (One of Six), CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0173 Insufficient evidence

How many whole fruits are inside this bowl in total, including any that are fully hidden?

Use only the photograph. A partly visible fruit counts as one fruit, but do not assume that every fruit must have some part visible from this viewpoint.

NC_A0042

Fumikas Sagisavas; File:Bowl of Spider Fruit.jpg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0174 Spatial relationships

Extend the straight upper boundary of the grey beam to the vertical line at90% of the image width. Where would that extension lie relative to the orange rock shape?

This is a line extrapolation on the drawing, not a claim that a laser passes through rock. Continue the same slope beyond the first rock contact. Use the orange polygon boundary rather than the caption text.

NC_A0016

Waikinl; File:Line of sight TLS1.png; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0175 Spatial relationships

Starting at the black dot, follow the black route through its first two complete straight segments. What is the net displacement in grid cells?

Count each grid cell as one unit. At a crossing keep following the same straight line until the next actual corner; do not turn merely because lines cross. Use image-right and image-down directions.

NC_A0061

Created by Ilmari Karonen; File:Knight's tour.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0176 Rotation and reflection

Reflect the long clock hand across the vertical line through the hub, then rotate that reflected hand90 degrees clockwise around the hub. In which image quadrant would its pointed tip finish?

Track the longer hand's far pointed tip, not the short counterweight behind the hub. Transform the hand only; keep the image directions and dial fixed. Reflection happens before rotation.

NC_A0062

Colin; File:Big Ben Inner Clock Face.jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant. Lossy image encoding was visually reviewed before testing.

VB0177 Rotation and reflection

First yaw the whole cube90 degrees clockwise as viewed from above. Then, looking straight at its new front face, roll the whole cube90 degrees clockwise. Which visible color from the original photograph finishes on top?

Define the original face on image-left as front, the face on image-right as right, and the upper face as top. After the yaw, front/right/top refer to fixed spatial positions, not original face identities. Turn the whole cube, not an individual layer.

NC_A0063

File:Rubiks cube solved.jpg: Mike Gonzalez (TheCoffee) derivative work: Pbroks13; File:Rubiks cube solved.png; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0178 Spatial relationships

Point A is one positive unit along the x-axis and point B one positive unit along the y-axis. Which colored sheet contains the entire straight segment from A to B?

The colored sheets are the coordinate planes. Use the axis labels in the image. A and B are not at the origin; do not mistake tinted overlaps for additional sheets.

NC_A0064

Sakurambo~commonswiki derivative work: Petrus3743; File:3D coordinate system.svg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0179 Spatial relationships

Under the coordinate convention shown, a rectangle has opposite corners(20,30) and(80,90). Translate the whole rectangle15 units toward image-right and10 units toward image-up. What are the new opposite corners?

Read the directions of increasing coordinates from this diagram. Both maximum coordinate values are at least100. Apply a translation, not a resize or reflection.

NC_A0065

Bert Niehaus; File:Graphics coordinate system.svg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0180 Sequential state tracking

A visual alarm activates only after two consecutive solar positions whose bright regions are partly covered by a foreground boat. By the final position in this composite, has the alarm activated?

Process all six solar positions from upper-left toward lower-right. The smooth Moon-shaped bite and the sea horizon are not boats. Only actual overlap of a boat silhouette with a bright solar region qualifies, not a boat elsewhere in the frame.

NC_A0013

Marika Caruana; File:Partial solar eclipse sequence.jpg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0181 Interface state decisions

The goal is Metallic at least0.80, Roughness at most0.40, and Alpha exactly1.00, leaving all other parameters unchanged. Which displayed parameters need changing?

Use the values currently displayed in this property panel. Changing one parameter does not automatically change the other two. Select the complete minimum set; no live application action is requested.

NUI_A0001

UnnecessaryDecision, CC0, via Wikimedia Commons (File:Blender 4.3.2 screenshot.png); credit: Own work, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0182 Charts and quantitative comparisons

In which panel do most points follow a broad curved rise-and-fall pattern that the blue straight line misses, rather than the main mismatch being one isolated outlier?

Judge the arrangement of most orange points. A single high or far-right point is not a broad curved pattern. Use panel positions, not an assumption that equal-looking blue lines imply equal data shapes.

NC_A0066

Anscombe.svg: Schutz (label using subscripts): Avenue; File:Anscombe's quartet 3.svg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0183 Charts and quantitative comparisons

Compared with the unshaded share of the middle pie, how does the unshaded share of the result pie on the right differ?

All circles represent the same whole. White is unshaded and grey is shaded. Use the radial partitions to read fractions; do not compare the total size of the circles on the page.

NC_A0067

Averater; File:Pie chart example 06.svg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0184 Charts and quantitative comparisons

Using the chart, how does the share of noncitizens in Gwinnett County compare with the share in Georgia?

The bars show citizenship, not noncitizenship. Treat the noncitizen share as100% minus the displayed citizenship share. Compare percentage points and choose the nearest offered magnitude. This concerns the pictured2015 chart, not current populations.

NC_A0068

Datawheel; File:Bar Chart of Citizenship in Gwinnett County, GA (2015).svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0185 Charts and quantitative comparisons

Which comparison correctly combines a higher median with a smaller interquartile range?

The thick line inside each box is its median. The box height, not the full whisker span, is the interquartile range. Read both properties before choosing.

NC_A0069

Jojotipe; File:Boxplot XYZ.svg; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0186 Causal and counterfactual reasoning

Replace the pictured load with a150N load while keeping the same pulley routing. What free-end pull would balance it in the ideal model?

Use the ratio between the displayed balanced load and the displayed free-end pull. For an unchanged ideal routing that ratio stays constant. Ignore friction and rope weight; do not count the hanging free end as an additional load-supporting strand.

NC_A0070

Polispasto2B.jpg: Mindbuilder derivative work: Arnaud Ramey (talk); File:Compound pulley.svg; CC BY-SA 3.0; via Wikimedia Commons, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0187 Causal and counterfactual reasoning

Starting from balance, double the mass on the longer arm. Which of these changes would restore balance while leaving both new masses fixed?

Use the mass-times-arm balance relation shown in the picture. In each option change only the named arm length; the other arm and the fulcrum position stay fixed. Choose among the offered actions, not every possible remedy.

NC_A0071

Tristanevanslee; File:Lever Principle 3D.svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0188 Causal and counterfactual reasoning

Close valve5 while leaving valve4 as drawn. On the next attempted downstroke of piston2, which of chambers1 and3 can receive displaced fluid through an open path from2?

For this diagram a vertical blue valve handle means closed and a diagonal handle means open. Closed valves completely block flow; open ones allow flow in either direction. Ignore leakage, fluid compression and any unpictured connections. Use the numbered chambers in the image.

NC_A0072

Wela49, Hydraulic press, CC BY-SA 3.0 via GFDL license migration, via Wikimedia Commons; canonical rasterization and metadata-free input conversion., CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0189 Causal and counterfactual reasoning

The red shaft turns clockwise. What are the blue shaft's directions before and after replacing the pictured belt routing with a belt that crosses once?

View both shafts from the same side. In this ideal model an uncrossed belt drives both shafts in the same direction; one crossing reverses their directions. Shafts stay fixed and the belt does not slip. Read the currently drawn routing for the before case.

NC_A0073

Pradana Aumars (talk · contribs); File:Belt pulley kinematic scheme.svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0190 Interface state decisions

Select a partition row whose file system is ext4, whose total Size is at most 100 GiB, and whose Unused space is at least 60 GiB. Which offered row satisfies all three requirements?

Use the table, not the relative widths of the colored blocks. This is a read-only selection question, not a request to change any disk. An extended container is not an ext4 file system.

NC_A0074

Pbmacsodak; File:Gparted screen showing partition.png; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0191 Interface state decisions

The goal is Always snap for objects, grids and guides, leaving all other settings unchanged. In which groups must the selected radio option change?

A blue filled circle marks the selected radio option. Each group permits one selected option. Choose the complete minimum set of groups that need changing from the pictured state.

NC_A0075

Stilfehler; File:Inkscape Document Properties Snap.png; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0192 Interface state decisions

Starting from the pictured state, export with run-length encoding and an alpha channel, without saving a comment. Which complete set of checkboxes needs toggling?

An empty checkbox is off. A comment is saved only when Save comment to file is on, even if the Comment text field contains text. Leave all unrelated options and field values unchanged.

NC_A0076

Gimp team; File:OpenGL Tutorial Gimp export as C.png; CC BY-SA 4.0; via Wikimedia Commons, CC BY-SA 4.0. Prepared model input from the cited source; see original source for its unmodified file.

VB0193 Charts and quantitative comparisons

Approximately what fraction of the states are represented by the first three histogram bins combined?

The vertical axis counts states, not land area. Add the counts in the first three adjacent bins, then compare with the total count across every nonempty bin. Choose the nearest offered percentage.

NC_A0080

HYanWong; File:R-US state areas-histogram.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0194 Charts and quantitative comparisons

Combining males and females, the population aged10 to14 is approximately how many times the population aged65 to69?

Use the pictured2022 population, not current estimates. Add the male and female bar magnitudes for each age band; the left side shows positive population magnitudes despite extending leftward. Choose the nearest offered ratio.

NC_A0081

Nosferattus; File:Belize Population Pyramid for 2022.svg; CC0; via Wikimedia Commons, CC0. Prepared model input from the cited source; see original source for its unmodified file.

VB0195 Causal and counterfactual reasoning

Disable the bottom-left transition. Starting with exactly the pictured tokens, what is the maximum number of times the bottom-right transition can fire?

Circles hold tokens and black bars are transitions. A firing consumes one token from every incoming arrow and produces one token on every outgoing arrow. Any enabled transition may fire repeatedly. Disabled transitions never fire. There are no hidden tokens or capacity limits.

NC_A0082

Msoos, Petri net example, CC BY-SA 2.5, via Wikimedia Commons; canonical rasterization and metadata-free input conversion., CC BY-SA 2.5. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0196 Causal and counterfactual reasoning

Force SPRINKLER to T externally. Under this intervention, what is the probability that GRASS WET is T?

Treat the diagram as the specified causal probability model. Forcing a node replaces its conditional probability table with the forced value and removes its incoming arrows; all other tables stay unchanged. Average over RAIN using its prior, rather than updating RAIN as if SPRINKLER had merely been observed. T and F mean true and false.

NC_A0083

AnAj; File:SimpleBayesNet.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0197 Navigation and planning

Just inside the upper-left entrance, choose between continuing down and turning right. Which choice can lead to the lower-right exit without returning through that first junction?

Black lines are impassable walls and white corridors are passable. The arrows mark entry and exit, not extra walls. You may follow bends and later junctions, but cannot cross a wall or leave the maze through its outside boundary.

NC_A0079

Vystrix Nexoth, Simple Maze, CC BY-SA 3.0 via GFDL license migration, via Wikimedia Commons; canonical rasterization and metadata-free input conversion., CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0198 Navigation and planning

Close the straight horizontal connection from the middle-left circle to the right-hand circle. What can be the two endpoints of a walk that uses every remaining connection exactly once?

Each drawn connection is an undirected edge, including each separate curved connection. You may revisit circles but cannot repeat or omit any remaining edge. The printed numbers describe the original graph, not the graph after closure. Either walking direction is allowed.

NC_A0084

Sarang; File:Königsberg Bridges graph.svg; Public domain; via Wikimedia Commons, Public domain. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.

VB0199 Navigation and planning

Start at a, visit every labeled node at least once, and return to a. What is the smallest possible total edge weight?

Connections are undirected. Add each traversed edge weight, including repeated traversals. Repeated nodes and edges are allowed; merely finding the shortest path to e does not meet the visit-all-and-return goal.

CCA0079

me, Public domain, via Wikimedia Commons (File:Graph weights.svg); credit: Own work, Public domain. Prepared model input from the cited source; see original source for its unmodified file.

VB0200 Navigation and planning

Choose a route from S to T that minimizes its largest single-edge weight, rather than its total weight. Which offered route achieves the best possible bottleneck?

Connections are undirected and each printed number is that edge's weight. A route's bottleneck is the maximum weight of any edge it uses. Smaller is better. Select a globally optimal route among the offered choices.

CCA0080

Willprice, CC BY-SA 3.0, via Wikimedia Commons (File:Dijkstra algorithm 1.svg); credit: Own work, CC BY-SA 3.0. Prepared model input from the cited source; see original source for its unmodified file. Re-encoded for the tested C1 input variant.