FA-99186 / Checkpoint tensor mapping / Open access

Checkpoint mapping: shard sizes are added along the wrong axis · case 01

shard sizes are added along the wrong axis.

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

ROOT CAUSE

The output always sums the leading dimension.

THE FAILURE

The output always sums the leading dimension.

Unsuccessful approach: Modulo indexing accepts negative axes outside the declared contract.

Case contract

Shapes have rank at least one and nonnegative extents. Return the concatenated shape when dimensions agree off axis; invalid axis or incompatible shapes return [].

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(shapes, axis):
    return [sum(s[i] for s in shapes) if i==0 else shapes[0][i] for i in range(len(shapes[0]))] if shapes else []
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('axis one', solve(*([[3, 2], [3, 3]], 1)), [3, 5])
check('axis zero', solve(*([[2, 3], [3, 3]], 0)), [5, 3])
check('singleton', solve(*([[3, 4]], 1)), [3, 4])
check('empty', solve(*([], 0)), [])
check('invalid', solve(*([[3]], -1)), [])
check('incompatible', solve(*([[3, 2], [4, 2]], 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
axis one	`[6, 2]`	`[3, 5]`	Failed
axis zero	`[5, 3]`	`[5, 3]`	Passed
singleton	`[3, 4]`	`[3, 4]`	Passed
empty	`[]`	`[]`	Passed
invalid	`[3]`	`[]`	Failed
incompatible	`[7, 2]`	`[]`	Failed

SHA-256 / 4e7ecce3dc3fa322363642875003046e74587bf431c4fecdd3fc66584da9555c

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(shapes, axis):
    return [sum(s[i] for s in shapes) if i==axis%len(shapes[0]) else shapes[0][i] for i in range(len(shapes[0]))] if shapes and len(shapes[0])>0 else []
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('axis one', solve(*([[3, 2], [3, 3]], 1)), [3, 5])
check('axis zero', solve(*([[2, 3], [3, 3]], 0)), [5, 3])
check('singleton', solve(*([[3, 4]], 1)), [3, 4])
check('empty', solve(*([], 0)), [])
check('invalid', solve(*([[3]], -1)), [])
check('incompatible', solve(*([[3, 2], [4, 2]], 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
axis one	`[3, 5]`	`[3, 5]`	Passed
axis zero	`[5, 3]`	`[5, 3]`	Passed
singleton	`[3, 4]`	`[3, 4]`	Passed
empty	`[]`	`[]`	Passed
invalid	`[3]`	`[]`	Failed
incompatible	`[3, 4]`	`[]`	Failed

SHA-256 / f89f00ba8455f519b9a6ae04264cbd0af643ef89e3911b9da4f1c3df09f4203e

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 2026-09-30T16:48:21.244491+00:00.

Case digest / 85686300b3833b1df6416e27fe9c1f614fad94ca949caaf39af11dcb4404bb49