110 gate-passed questions · 36 cells · 19 occupied · go to this type's cells
Bins were cut on this corpus at 240×135@16 (CLEAR_PX 25, motion floor 19 px). Margin bins: a near-edge band [1, 1.25) above the gate boundary, then corpus quantiles; salience bins: a visibility floor, then quantiles.
cover_max / 0.15; 'neither': rect_gap / 0.5 · measurable on 110 questions
| margin bin | numeric range |
|---|---|
| near-edge [1,1.25) | [−∞, 1.25) |
| bin0 | [1.25, 3.057) |
| bin1 | [3.057, ∞) |
Ask: is this the same quantity the gate certifies on for this type? Are the states (e.g. 'same', 'gone') the right one-bin exceptions? Are the edges where a change in difficulty actually happens?
min queried-object peak area (frac of frame) · floor 0.0007716
| salience bin | numeric range |
|---|---|
| <floor (0.0008) | [−∞, 0.0007716) |
| bin0 | [0.0007716, 0.009444) |
| bin1 | [0.009444, 0.01586) |
| bin2 | [0.01586, ∞) |
Ask: does this axis capture what makes a question of this type perceptually hard, independently of the answer? Is the floor the right anti-gaming boundary?
property of the correct answer (branch 2): occluder's role ('neither' when none) · labels: mover, neither, static
Compatible margin bins per context label (cells outside these are definitionally impossible):
| answer | n |
|---|---|
| neither: they never overlap on screen | 74 |
| the green cylinder | 4 |
| the green cone | 3 |
| the red cone | 3 |
| the orange ball | 2 |
| the purple cylinder | 2 |
| the purple cube | 2 |
| the red cylinder | 2 |
| the blue cylinder | 2 |
| the green cube | 2 |
| the blue-and-orange striped cube | 2 |
| the blue cone | 1 |
| the purple cone | 1 |
| the red-and-yellow striped ball | 1 |
| the yellow cube | 1 |
| the orange cube | 1 |
| the green-and-yellow striped cone | 1 |
| the blue-and-green striped cube | 1 |
| the green-and-orange striped cube | 1 |
| the blue-and-purple striped cube | 1 |
| the purple-and-orange striped cube | 1 |
| the yellow cylinder | 1 |
| the red cube | 1 |