FA-100071 / Preference pair construction / Open access

Preference pairs: sub-threshold rating gaps are accepted · case 01

sub-threshold rating gaps are accepted.

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

ROOT CAUSE

Any positive rating difference is accepted even when the task requires a minimum margin.

THE FAILURE

Any positive rating difference is accepted even when the task requires a minimum margin.

Unsuccessful approach: Using a strict greater-than comparison incorrectly rejects a gap exactly at the margin.

Case contract

Rows are [chosen,rejected,chosen_score,rejected_score]. Return endpoints only when the gap is at least margin.

Why this case matters

A small deterministic preference-data transform makes pair semantics inspectable without a training framework.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(rows, margin):
    return [[r[0],r[1]] for r in rows if r[2]>r[3]]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('below margin',solve([['a','b',N+1,N]],2),[])
check('exact margin',solve([['a','b',N+2,N]],2),[['a','b']])
check('above margin',solve([['a','b',N+3,N]],2),[['a','b']])
check('tie',solve([['a','b',N,N]],1),[])
check('empty',solve([],N),[])
check('zero margin strict preference',solve([['a','b',1,0]],0),[['a','b']])
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
below margin	`[['a', 'b']]`	`[]`	Failed
exact margin	`[['a', 'b']]`	`[['a', 'b']]`	Passed
above margin	`[['a', 'b']]`	`[['a', 'b']]`	Passed
tie	`[]`	`[]`	Passed
empty	`[]`	`[]`	Passed
zero margin strict preference	`[['a', 'b']]`	`[['a', 'b']]`	Passed

SHA-256 / 137330ef27d97a5475393454c0dfde8ce6aa8a8e2c84f8a7cd4b7f7a2e38ed0d

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(rows, margin):
    return [[r[0],r[1]] for r in rows if r[2]-r[3]>margin]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('below margin',solve([['a','b',N+1,N]],2),[])
check('exact margin',solve([['a','b',N+2,N]],2),[['a','b']])
check('above margin',solve([['a','b',N+3,N]],2),[['a','b']])
check('tie',solve([['a','b',N,N]],1),[])
check('empty',solve([],N),[])
check('zero margin strict preference',solve([['a','b',1,0]],0),[['a','b']])
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
below margin	`[]`	`[]`	Passed
exact margin	`[]`	`[['a', 'b']]`	Failed
above margin	`[['a', 'b']]`	`[['a', 'b']]`	Passed
tie	`[]`	`[]`	Passed
empty	`[]`	`[]`	Passed
zero margin strict preference	`[['a', 'b']]`	`[['a', 'b']]`	Passed

SHA-256 / 21df564e0da17b152731fd3c8f72489ffcfe05bf85df81f0f67cb260319161c1

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

These are explicit toy record contracts; they do not claim compatibility with a trainer or improved model behavior. 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 2026-09-30T16:48:31.261611+00:00.

Case digest / 89d32634eb69c9378462b947cf32419c69cc8cc1283ef1316be56c4f9806820b