FA-98136 / Sequence batch packing / Open access

Padding positions inherit the last segment owner · case 01

Padding positions inherit the last segment owner.

Verified by executionVariant 1 · 6 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Padding repeats the last valid owner, making padding indistinguishable from sequence content.

THE FAILURE

Padding repeats the last valid owner, making padding indistinguishable from sequence content.

Unsuccessful approach: A max guard masks overfull rows instead of enforcing the stated width precondition.

Case contract

Expand segment lengths into owner IDs and pad to width with sentinel -1; total length must not exceed width.

Why this case matters

This deterministic toy packing rule models a common boundary in preparing bounded training sequences.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(lengths, width):
    owners=[i for i,n in enumerate(lengths) for _ in range(n)]; return owners+[owners[-1] if owners else -1]*(width-len(owners))
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('nonempty padding', solve([N,1],N+3), [0]*N+[1,-1,-1])
check('exact width', solve([N],N), [0]*N)
check('empty row', solve([],N), [-1]*N)
check('zero segment', solve([0,N],N), [1]*N)
check('one valid slot', solve([1],3), [0,-1,-1])
check('mixed owners', solve([1,2],5), [0,1,1,-1,-1])
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 fixture	Actual	Expected	Outcome
nonempty padding	`[0, 1, 1, 1]`	`[0, 1, -1, -1]`	Failed
exact width	`[0]`	`[0]`	Passed
empty row	`[-1]`	`[-1]`	Passed
zero segment	`[1]`	`[1]`	Passed
one valid slot	`[0, 0, 0]`	`[0, -1, -1]`	Failed
mixed owners	`[0, 1, 1, 1, 1]`	`[0, 1, 1, -1, -1]`	Failed

SHA-256 / 039c7f2be08cd742d39079caf8bc60bf2b46cf1116ae238b48da3128260b0c82

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(lengths, width):
    owners=[i for i,n in enumerate(lengths) for _ in range(n)]; return owners+[-1]*max(0,width-len(owners)-1)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('nonempty padding', solve([N,1],N+3), [0]*N+[1,-1,-1])
check('exact width', solve([N],N), [0]*N)
check('empty row', solve([],N), [-1]*N)
check('zero segment', solve([0,N],N), [1]*N)
check('one valid slot', solve([1],3), [0,-1,-1])
check('mixed owners', solve([1,2],5), [0,1,1,-1,-1])
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 fixture	Actual	Expected	Outcome
nonempty padding	`[0, 1, -1]`	`[0, 1, -1, -1]`	Failed
exact width	`[0]`	`[0]`	Passed
empty row	`[]`	`[-1]`	Failed
zero segment	`[1]`	`[1]`	Passed
one valid slot	`[0, -1]`	`[0, -1, -1]`	Failed
mixed owners	`[0, 1, 1, -1]`	`[0, 1, 1, -1, -1]`	Failed

SHA-256 / c582c0bc9635118df56794e69908eba7e6734c8c78f29aab880f6488173ec577

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 6 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Verification & scope

Inputs are short integer-token lists under the stated toy contract; results do not claim compatibility with a specific training framework. 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 2026-09-30T16:48:09.758170+00:00.

Case digest / 705868e8de14211d7fb3aee4e4719da9bc511a40b1663eeda51c1a766e3fe3cc