FAILURE MAP
← Case archive

FA-86671 / Procedural level generation constraints / Open access

Wall autotile bitmask: Mask keeps only the last neighbour · case 01

Wall corners use straight-wall sprites.

Verified by executionVariant 1 · 8 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Each neighbour overwrites the mask instead of or-ing its bit.

VERIFIED REPAIR

Restore `m |= bit` at the mask accumulation step.

Unsuccessful approach: Keeping the largest bit still loses combinations.

Case contract

For each '#' cell the mask sums N=1, E=2, S=4, W=8 for 4-neighbours that are '#' (row 0 is north). Out-of-bounds neighbours count as walls when same_oob. Non-wall cells get -1. Returns rows of integers.

Why this case matters

Procedural generators silently emit unplayable or unfair levels when a single constraint check uses the wrong boundary, axis, neighborhood or update order; the defect is visible in exact generated geometry.

1 / The failure

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(grid, same_oob):
    h = len(grid)
    w = len(grid[0])
    out = []
    for r in range(h):
        row = []
        for c in range(w):
            if grid[r][c] != '#':
                row.append(-1)
                continue
            m = 0
            for bit, (dr, dc) in ((1, (-1, 0)), (2, (0, 1)), (4, (1, 0)), (8, (0, -1))):
                rr, cc = r + dr, c + dc
                if 0 <= rr < h and 0 <= cc < w:
                    if grid[rr][cc] == '#':
                        m = bit
                elif same_oob:
                    m |= bit
            row.append(m)
        out.append(row)
    return out
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
cases = [[('regression mask accumulation #1',
   [['~###', '##~.', 'D~~.', '#D#.', '#..#'], True],
   [[-1, 7, 11, 11], [10, 9, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]]),
  ('regression mask accumulation #2',
   [['##D', '..#', '##D'], True],
   [[11, 9, -1], [-1, -1, 2], [14, 12, -1]]),
  ('regression mask accumulation #3', [['####D'], False], [[2, 10, 10, 8, -1]]),
  ('regression mask accumulation #4',
   [['##.D', '.D#~', '##..', '#.#~', '~D.#'], True],
   [[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('control #1', [['.#D', '#.~'], True], [[-1, 1, -1], [12, -1, -1]])],
 [('regression mask accumulation #1', [['####D'], False], [[2, 10, 10, 8, -1]]),
  ('regression mask accumulation #2',
   [['##.D', '.D#~', '##..', '#.#~', '~D.#'], True],
   [[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]),
  ('regression mask accumulation #3',
   [['#..D.', '.##.#', '..##D', 'D.#.D'], False],
   [[0, -1, -1, -1, -1], [-1, 2, 12, -1, 0], [-1, -1, 7, 8, -1], [-1, -1, 1, -1, -1]]),
  ('regression mask accumulation #4', [['~.#', '##D'], True], [[-1, -1, 3], [14, 12, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('control #1', [['.#D', '#.~'], True], [[-1, 1, -1], [12, -1, -1]]),
  ('control #2', [['D~#', '#~.'], False], [[-1, -1, 0], [0, -1, -1]])],
 [('regression mask accumulation #1',
   [['#..D.', '.##.#', '..##D', 'D.#.D'], False],
   [[0, -1, -1, -1, -1], [-1, 2, 12, -1, 0], [-1, -1, 7, 8, -1], [-1, -1, 1, -1, -1]]),
  ('regression mask accumulation #2', [['~.#', '##D'], True], [[-1, -1, 3], [14, 12, -1]]),
  ('regression mask accumulation #3', [['#', '#', '#'], False], [[4], [5], [1]]),
  ('regression mask accumulation #4',
   [['~#.~', '.D~#', '###~', '#~#D', '.##.'], False],
   [[-1, 0, -1, -1], [-1, -1, -1, 0], [6, 10, 12, -1], [1, -1, 5, -1], [-1, 2, 9, -1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('control #1', [['D~#', '#~.'], False], [[-1, -1, 0], [0, -1, -1]]),
  ('control #2', [['~#', '.~', '.#'], True], [[-1, 3], [-1, -1], [-1, 6]])],
 [('regression mask accumulation #1', [['#', '#', '#'], False], [[4], [5], [1]]),
  ('regression mask accumulation #2',
   [['~#.~', '.D~#', '###~', '#~#D', '.##.'], False],
   [[-1, 0, -1, -1], [-1, -1, -1, 0], [6, 10, 12, -1], [1, -1, 5, -1], [-1, 2, 9, -1]]),
  ('regression mask accumulation #3',
   [['D#DD', '#D.~', '###~', '#~##'], False],
   [[-1, 0, -1, -1], [4, -1, -1, -1], [7, 10, 12, -1], [1, -1, 3, 8]]),
  ('regression mask accumulation #4',
   [['#D.#.', 'D##~#', 'DD#~#'], False],
   [[0, -1, -1, 0, -1], [-1, 2, 12, -1, 4], [-1, -1, 1, -1, 1]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('control #1', [['~', '~'], True], [[-1], [-1]]),
  ('control #2',
   [['#~~', '..D', '~#.', 'D.D'], False],
   [[0, -1, -1], [-1, -1, -1], [-1, 0, -1], [-1, -1, -1]])],
 [('regression mask accumulation #1',
   [['D#DD', '#D.~', '###~', '#~##'], False],
   [[-1, 0, -1, -1], [4, -1, -1, -1], [7, 10, 12, -1], [1, -1, 3, 8]]),
  ('regression mask accumulation #2',
   [['#D.#.', 'D##~#', 'DD#~#'], False],
   [[0, -1, -1, 0, -1], [-1, 2, 12, -1, 4], [-1, -1, 1, -1, 1]]),
  ('regression mask accumulation #3', [['###', '.##'], True], [[11, 15, 15], [-1, 7, 15]]),
  ('regression mask accumulation #4', [['~D~#', '####'], True], [[-1, -1, -1, 7], [14, 14, 14, 15]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('control #1', [['.'], True], [[-1]]),
  ('control #2', [['~D.D#', '#~.#D'], True], [[-1, -1, -1, -1, 3], [12, -1, -1, 4, -1]])]]
for label, args, expected in cases[N-1]:
    check(label, solve(*args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression mask accumulation #1[[-1, 4, 8, 8], [10, 8, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]][[-1, 7, 11, 11], [10, 9, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]]Failed
regression mask accumulation #2[[10, 8, -1], [-1, -1, 2], [14, 8, -1]][[11, 9, -1], [-1, -1, 2], [14, 12, -1]]Failed
regression mask accumulation #3[[2, 8, 8, 8, -1]][[2, 10, 10, 8, -1]]Failed
regression mask accumulation #4[[10, 8, -1, -1], [-1, -1, 0, -1], [12, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]][[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]Failed
vertical wall pair #1[[4], [1]][[4], [1]]Passed
door next to wall #1[[0, -1]][[0, -1]]Passed
lone wall at edge #1[[15]][[15]]Passed
control #1[[-1, 1, -1], [12, -1, -1]][[-1, 1, -1], [12, -1, -1]]Passed

SHA-256 / 1d392dabab6a7f178d06d6c984600d16cf14aec849ad0f43a6ffb59cd8a9c888

2 / The unsuccessful fix

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(grid, same_oob):
    h = len(grid)
    w = len(grid[0])
    out = []
    for r in range(h):
        row = []
        for c in range(w):
            if grid[r][c] != '#':
                row.append(-1)
                continue
            m = 0
            for bit, (dr, dc) in ((1, (-1, 0)), (2, (0, 1)), (4, (1, 0)), (8, (0, -1))):
                rr, cc = r + dr, c + dc
                if 0 <= rr < h and 0 <= cc < w:
                    if grid[rr][cc] == '#':
                        m = max(m, bit)
                elif same_oob:
                    m |= bit
            row.append(m)
        out.append(row)
    return out
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
cases = [[('regression mask accumulation #1',
   [['~###', '##~.', 'D~~.', '#D#.', '#..#'], True],
   [[-1, 7, 11, 11], [10, 9, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]]),
  ('regression mask accumulation #2',
   [['##D', '..#', '##D'], True],
   [[11, 9, -1], [-1, -1, 2], [14, 12, -1]]),
  ('regression mask accumulation #3', [['####D'], False], [[2, 10, 10, 8, -1]]),
  ('regression mask accumulation #4',
   [['##.D', '.D#~', '##..', '#.#~', '~D.#'], True],
   [[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('control #1', [['.#D', '#.~'], True], [[-1, 1, -1], [12, -1, -1]])],
 [('regression mask accumulation #1', [['####D'], False], [[2, 10, 10, 8, -1]]),
  ('regression mask accumulation #2',
   [['##.D', '.D#~', '##..', '#.#~', '~D.#'], True],
   [[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]),
  ('regression mask accumulation #3',
   [['#..D.', '.##.#', '..##D', 'D.#.D'], False],
   [[0, -1, -1, -1, -1], [-1, 2, 12, -1, 0], [-1, -1, 7, 8, -1], [-1, -1, 1, -1, -1]]),
  ('regression mask accumulation #4', [['~.#', '##D'], True], [[-1, -1, 3], [14, 12, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('control #1', [['.#D', '#.~'], True], [[-1, 1, -1], [12, -1, -1]]),
  ('control #2', [['D~#', '#~.'], False], [[-1, -1, 0], [0, -1, -1]])],
 [('regression mask accumulation #1',
   [['#..D.', '.##.#', '..##D', 'D.#.D'], False],
   [[0, -1, -1, -1, -1], [-1, 2, 12, -1, 0], [-1, -1, 7, 8, -1], [-1, -1, 1, -1, -1]]),
  ('regression mask accumulation #2', [['~.#', '##D'], True], [[-1, -1, 3], [14, 12, -1]]),
  ('regression mask accumulation #3', [['#', '#', '#'], False], [[4], [5], [1]]),
  ('regression mask accumulation #4',
   [['~#.~', '.D~#', '###~', '#~#D', '.##.'], False],
   [[-1, 0, -1, -1], [-1, -1, -1, 0], [6, 10, 12, -1], [1, -1, 5, -1], [-1, 2, 9, -1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('control #1', [['D~#', '#~.'], False], [[-1, -1, 0], [0, -1, -1]]),
  ('control #2', [['~#', '.~', '.#'], True], [[-1, 3], [-1, -1], [-1, 6]])],
 [('regression mask accumulation #1', [['#', '#', '#'], False], [[4], [5], [1]]),
  ('regression mask accumulation #2',
   [['~#.~', '.D~#', '###~', '#~#D', '.##.'], False],
   [[-1, 0, -1, -1], [-1, -1, -1, 0], [6, 10, 12, -1], [1, -1, 5, -1], [-1, 2, 9, -1]]),
  ('regression mask accumulation #3',
   [['D#DD', '#D.~', '###~', '#~##'], False],
   [[-1, 0, -1, -1], [4, -1, -1, -1], [7, 10, 12, -1], [1, -1, 3, 8]]),
  ('regression mask accumulation #4',
   [['#D.#.', 'D##~#', 'DD#~#'], False],
   [[0, -1, -1, 0, -1], [-1, 2, 12, -1, 4], [-1, -1, 1, -1, 1]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('control #1', [['~', '~'], True], [[-1], [-1]]),
  ('control #2',
   [['#~~', '..D', '~#.', 'D.D'], False],
   [[0, -1, -1], [-1, -1, -1], [-1, 0, -1], [-1, -1, -1]])],
 [('regression mask accumulation #1',
   [['D#DD', '#D.~', '###~', '#~##'], False],
   [[-1, 0, -1, -1], [4, -1, -1, -1], [7, 10, 12, -1], [1, -1, 3, 8]]),
  ('regression mask accumulation #2',
   [['#D.#.', 'D##~#', 'DD#~#'], False],
   [[0, -1, -1, 0, -1], [-1, 2, 12, -1, 4], [-1, -1, 1, -1, 1]]),
  ('regression mask accumulation #3', [['###', '.##'], True], [[11, 15, 15], [-1, 7, 15]]),
  ('regression mask accumulation #4', [['~D~#', '####'], True], [[-1, -1, -1, 7], [14, 14, 14, 15]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('control #1', [['.'], True], [[-1]]),
  ('control #2', [['~D.D#', '#~.#D'], True], [[-1, -1, -1, -1, 3], [12, -1, -1, 4, -1]])]]
for label, args, expected in cases[N-1]:
    check(label, solve(*args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression mask accumulation #1[[-1, 4, 8, 8], [10, 8, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]][[-1, 7, 11, 11], [10, 9, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]]Failed
regression mask accumulation #2[[10, 8, -1], [-1, -1, 2], [14, 8, -1]][[11, 9, -1], [-1, -1, 2], [14, 12, -1]]Failed
regression mask accumulation #3[[2, 8, 8, 8, -1]][[2, 10, 10, 8, -1]]Failed
regression mask accumulation #4[[10, 8, -1, -1], [-1, -1, 0, -1], [12, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]][[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]Failed
vertical wall pair #1[[4], [1]][[4], [1]]Passed
door next to wall #1[[0, -1]][[0, -1]]Passed
lone wall at edge #1[[15]][[15]]Passed
control #1[[-1, 1, -1], [12, -1, -1]][[-1, 1, -1], [12, -1, -1]]Passed

SHA-256 / ceb3b86a964f0e79ca2c276c656505e2819a0589ab6a918a72b5317bfc4f9726

3 / The verified repair

Exit 0
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(grid, same_oob):
    h = len(grid)
    w = len(grid[0])
    out = []
    for r in range(h):
        row = []
        for c in range(w):
            if grid[r][c] != '#':
                row.append(-1)
                continue
            m = 0
            for bit, (dr, dc) in ((1, (-1, 0)), (2, (0, 1)), (4, (1, 0)), (8, (0, -1))):
                rr, cc = r + dr, c + dc
                if 0 <= rr < h and 0 <= cc < w:
                    if grid[rr][cc] == '#':
                        m |= bit
                elif same_oob:
                    m |= bit
            row.append(m)
        out.append(row)
    return out
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
cases = [[('regression mask accumulation #1',
   [['~###', '##~.', 'D~~.', '#D#.', '#..#'], True],
   [[-1, 7, 11, 11], [10, 9, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]]),
  ('regression mask accumulation #2',
   [['##D', '..#', '##D'], True],
   [[11, 9, -1], [-1, -1, 2], [14, 12, -1]]),
  ('regression mask accumulation #3', [['####D'], False], [[2, 10, 10, 8, -1]]),
  ('regression mask accumulation #4',
   [['##.D', '.D#~', '##..', '#.#~', '~D.#'], True],
   [[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('control #1', [['.#D', '#.~'], True], [[-1, 1, -1], [12, -1, -1]])],
 [('regression mask accumulation #1', [['####D'], False], [[2, 10, 10, 8, -1]]),
  ('regression mask accumulation #2',
   [['##.D', '.D#~', '##..', '#.#~', '~D.#'], True],
   [[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]),
  ('regression mask accumulation #3',
   [['#..D.', '.##.#', '..##D', 'D.#.D'], False],
   [[0, -1, -1, -1, -1], [-1, 2, 12, -1, 0], [-1, -1, 7, 8, -1], [-1, -1, 1, -1, -1]]),
  ('regression mask accumulation #4', [['~.#', '##D'], True], [[-1, -1, 3], [14, 12, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('control #1', [['.#D', '#.~'], True], [[-1, 1, -1], [12, -1, -1]]),
  ('control #2', [['D~#', '#~.'], False], [[-1, -1, 0], [0, -1, -1]])],
 [('regression mask accumulation #1',
   [['#..D.', '.##.#', '..##D', 'D.#.D'], False],
   [[0, -1, -1, -1, -1], [-1, 2, 12, -1, 0], [-1, -1, 7, 8, -1], [-1, -1, 1, -1, -1]]),
  ('regression mask accumulation #2', [['~.#', '##D'], True], [[-1, -1, 3], [14, 12, -1]]),
  ('regression mask accumulation #3', [['#', '#', '#'], False], [[4], [5], [1]]),
  ('regression mask accumulation #4',
   [['~#.~', '.D~#', '###~', '#~#D', '.##.'], False],
   [[-1, 0, -1, -1], [-1, -1, -1, 0], [6, 10, 12, -1], [1, -1, 5, -1], [-1, 2, 9, -1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('control #1', [['D~#', '#~.'], False], [[-1, -1, 0], [0, -1, -1]]),
  ('control #2', [['~#', '.~', '.#'], True], [[-1, 3], [-1, -1], [-1, 6]])],
 [('regression mask accumulation #1', [['#', '#', '#'], False], [[4], [5], [1]]),
  ('regression mask accumulation #2',
   [['~#.~', '.D~#', '###~', '#~#D', '.##.'], False],
   [[-1, 0, -1, -1], [-1, -1, -1, 0], [6, 10, 12, -1], [1, -1, 5, -1], [-1, 2, 9, -1]]),
  ('regression mask accumulation #3',
   [['D#DD', '#D.~', '###~', '#~##'], False],
   [[-1, 0, -1, -1], [4, -1, -1, -1], [7, 10, 12, -1], [1, -1, 3, 8]]),
  ('regression mask accumulation #4',
   [['#D.#.', 'D##~#', 'DD#~#'], False],
   [[0, -1, -1, 0, -1], [-1, 2, 12, -1, 4], [-1, -1, 1, -1, 1]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('door next to wall #1', [['#D'], False], [[0, -1]]),
  ('control #1', [['~', '~'], True], [[-1], [-1]]),
  ('control #2',
   [['#~~', '..D', '~#.', 'D.D'], False],
   [[0, -1, -1], [-1, -1, -1], [-1, 0, -1], [-1, -1, -1]])],
 [('regression mask accumulation #1',
   [['D#DD', '#D.~', '###~', '#~##'], False],
   [[-1, 0, -1, -1], [4, -1, -1, -1], [7, 10, 12, -1], [1, -1, 3, 8]]),
  ('regression mask accumulation #2',
   [['#D.#.', 'D##~#', 'DD#~#'], False],
   [[0, -1, -1, 0, -1], [-1, 2, 12, -1, 4], [-1, -1, 1, -1, 1]]),
  ('regression mask accumulation #3', [['###', '.##'], True], [[11, 15, 15], [-1, 7, 15]]),
  ('regression mask accumulation #4', [['~D~#', '####'], True], [[-1, -1, -1, 7], [14, 14, 14, 15]]),
  ('lone wall at edge #1', [['#'], True], [[15]]),
  ('vertical wall pair #1', [['#', '#'], False], [[4], [1]]),
  ('control #1', [['.'], True], [[-1]]),
  ('control #2', [['~D.D#', '#~.#D'], True], [[-1, -1, -1, -1, 3], [12, -1, -1, 4, -1]])]]
for label, args, expected in cases[N-1]:
    check(label, solve(*args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression mask accumulation #1[[-1, 7, 11, 11], [10, 9, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]][[-1, 7, 11, 11], [10, 9, -1, -1], [-1, -1, -1, -1], [12, -1, 0, -1], [13, -1, -1, 6]]Passed
regression mask accumulation #2[[11, 9, -1], [-1, -1, 2], [14, 12, -1]][[11, 9, -1], [-1, -1, 2], [14, 12, -1]]Passed
regression mask accumulation #3[[2, 10, 10, 8, -1]][[2, 10, 10, 8, -1]]Passed
regression mask accumulation #4[[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]][[11, 9, -1, -1], [-1, -1, 0, -1], [14, 8, -1, -1], [9, -1, 0, -1], [-1, -1, -1, 6]]Passed
vertical wall pair #1[[4], [1]][[4], [1]]Passed
door next to wall #1[[0, -1]][[0, -1]]Passed
lone wall at edge #1[[15]][[15]]Passed
control #1[[-1, 1, -1], [12, -1, -1]][[-1, 1, -1], [12, -1, -1]]Passed

SHA-256 / 9f44fd75f8d9dc9f79dc1c62a5ef1add09a7f2e329c429803730a123acfc7332

Verification & scope

Deterministic toy contract stipulated for this model; integer or exact arithmetic only, not a reproduction of any specific game engine. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:50:51.773307+00:00.

Case digest / 5c66478a1bb2eb946a824044c973a264f5f35e3bd84148b52bd23d327fe9dfdd