FA-98816 / Tensor stride and reshape semantics / Open access

Tensor views: contiguity ignores strides on unit axes · case 01

contiguity ignores strides on unit axes.

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

ROOT CAUSE

Contiguity depends on stride progression for non-singleton axes, not every stored stride value.

THE FAILURE

Contiguity depends on stride progression for non-singleton axes, not every stored stride value.

Unsuccessful approach: Requiring progression through singleton axes rejects layouts whose addresses are unaffected by those strides.

Case contract

Dense row-major contiguity checks expected stride only for dimensions larger than one.

Why this case matters

Dense tensor views rely on consistent logical coordinates and physical storage addresses.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(shape, strides):
    return all(s==1 for s in strides)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('contiguous',solve([2,3],[3,1]),True)
check('transposed',solve([2,3],[1,2]),False)
check('unit arbitrary',solve([1,3],[99,1]),True)
check('scalar',solve([],[]),True)
check('all singleton',solve([1,1],[7,0]),True)
check('gap',solve([2,3],[4,1]),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
contiguous	`False`	`True`	Failed
transposed	`False`	`False`	Passed
unit arbitrary	`False`	`True`	Failed
scalar	`True`	`True`	Passed
all singleton	`False`	`True`	Failed
gap	`False`	`False`	Passed

SHA-256 / 97833b8004d69dffe3c595aafe1814c5a42d925955f9c5fde5d31906bd0be521

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(shape, strides):
    expected=1
    for d,s in zip(reversed(shape),reversed(strides)):
     if s!=expected: return False
     expected*=d
    return True
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('contiguous',solve([2,3],[3,1]),True)
check('transposed',solve([2,3],[1,2]),False)
check('unit arbitrary',solve([1,3],[99,1]),True)
check('scalar',solve([],[]),True)
check('all singleton',solve([1,1],[7,0]),True)
check('gap',solve([2,3],[4,1]),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
contiguous	`True`	`True`	Passed
transposed	`False`	`False`	Passed
unit arbitrary	`False`	`True`	Failed
scalar	`True`	`True`	Passed
all singleton	`False`	`True`	Failed
gap	`False`	`False`	Passed

SHA-256 / ebccc05e0f3f941997a11bf7b5e14119956757d2580c6064e738ce795fd3a482

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

Self-contained integer metadata models only; no production framework compatibility or performance claim. 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 2026-09-30T16:48:17.165394+00:00.

Case digest / a509739eadea5ec57dc9bfa3e6012042bad15c23a53c5c06119db041b8ffb757