FA-99151 / Checkpoint tensor mapping / Open access

Checkpoint mapping: a wildcard axis masks mismatches in fixed axes · case 01

a wildcard axis masks mismatches in fixed axes.

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

ROOT CAUSE

Wildcard presence anywhere bypasses all fixed-axis comparisons.

THE FAILURE

Wildcard presence anywhere bypasses all fixed-axis comparisons.

Unsuccessful approach: Allowing an actual -1 treats metadata as a template and accepts an unresolved shape.

Case contract

A template dimension -1 accepts any nonnegative extent at that position; other axes must match.

Why this case matters

A bounded offline model of checkpoint tensor bookkeeping in a training workflow.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(actual, template):
    return len(actual)==len(template) and all(a==b or -1 in template for a,b in zip(actual,template))
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('declared', solve(*([3, 4], [3, -1])), True)
check('wrong fixed', solve(*([4, 4], [3, -1])), False)
check('actual wildcard', solve(*([-1, 4], [3, -1])), False)
check('rank', solve(*([3], [3, -1])), False)
check('empty', solve(*([], [])), True)
check('negative at wildcard', solve(*([3, -1], [3, -1])), False)
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
declared	`True`	`True`	Passed
wrong fixed	`True`	`False`	Failed
actual wildcard	`True`	`False`	Failed
rank	`False`	`False`	Passed
empty	`True`	`True`	Passed
negative at wildcard	`True`	`False`	Failed

SHA-256 / c4d1630d8288edcfebffea08ebb43430f5f68e364dcc28015b4deca379b7c699

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(actual, template):
    return len(actual)==len(template) and all(b==-1 or a==b or a==-1 for a,b in zip(actual,template))
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('declared', solve(*([3, 4], [3, -1])), True)
check('wrong fixed', solve(*([4, 4], [3, -1])), False)
check('actual wildcard', solve(*([-1, 4], [3, -1])), False)
check('rank', solve(*([3], [3, -1])), False)
check('empty', solve(*([], [])), True)
check('negative at wildcard', solve(*([3, -1], [3, -1])), False)
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
declared	`True`	`True`	Passed
wrong fixed	`False`	`False`	Passed
actual wildcard	`True`	`False`	Failed
rank	`False`	`False`	Passed
empty	`True`	`True`	Passed
negative at wildcard	`True`	`False`	Failed

SHA-256 / 4e88a33be6aecaf0436183e634898ced9725d450d9dd52a7bae9714b6ede4be9

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 already decoded Python values; this toy contract makes no claim about a framework, file format, or production checkpoint. 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 2026-09-30T16:48:20.799442+00:00.

Case digest / 91bc99be8e9be25eec85793f39591f43009663b74986964821b5b3f61fa2355c