FA-100096 / Preference pair construction / Open access

Preference pairs: pair cap is applied before duplicate removal · case 01

pair cap is applied before duplicate removal.

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

ROOT CAUSE

Repeated pair keys consume output capacity and hide later unique training pairs.

THE FAILURE

Repeated pair keys consume output capacity and hide later unique training pairs.

Unsuccessful approach: Deduplicating by chosen response alone merges different rejected alternatives.

Case contract

Rows are [pair_key,chosen,rejected]. Keep the first row per pair_key, then return the first limit rows; limit is nonnegative.

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, limit):
    return rows[:limit]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
a='a'+str(N); b='b'+str(N); c='c'+str(N)
rows=[[a,'x','y'],[a,'x','z'],[b,'u','v'],[c,'m','n']]
check('dedup before cap',solve(rows,2),[[a,'x','y'],[b,'u','v']])
check('zero cap',solve(rows,0),[])
check('empty rows',solve([],N),[])
check('cap beyond size',solve(rows,9),[[a,'x','y'],[b,'u','v'],[c,'m','n']])
check('distinct rejected alternatives',solve([[a,'x','y'],[b,'x','z']],3),[[a,'x','y'],[b,'x','z']])
check('one row',solve([[a,'x','y']],1),[[a,'x','y']])
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
dedup before cap	`[['a1', 'x', 'y'], ['a1', 'x', 'z']]`	`[['a1', 'x', 'y'], ['b1', 'u', 'v']]`	Failed
zero cap	`[]`	`[]`	Passed
empty rows	`[]`	`[]`	Passed
cap beyond size	`[['a1', 'x', 'y'], ['a1', 'x', 'z'], ['b1', 'u', 'v'], ['c1', 'm', 'n']]`	`[['a1', 'x', 'y'], ['b1', 'u', 'v'], ['c1', 'm', 'n']]`	Failed
distinct rejected alternatives	`[['a1', 'x', 'y'], ['b1', 'x', 'z']]`	`[['a1', 'x', 'y'], ['b1', 'x', 'z']]`	Passed
one row	`[['a1', 'x', 'y']]`	`[['a1', 'x', 'y']]`	Passed

SHA-256 / c78e3b7e8ceb5f08ad64430f62b85a5dd9572fd9f4a4f8c4d690972fb2802f09

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(rows, limit):
    seen=set(); out=[]
    for r in rows[:limit]:
     if r[1] not in seen: seen.add(r[1]); out.append(r)
    return out
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
a='a'+str(N); b='b'+str(N); c='c'+str(N)
rows=[[a,'x','y'],[a,'x','z'],[b,'u','v'],[c,'m','n']]
check('dedup before cap',solve(rows,2),[[a,'x','y'],[b,'u','v']])
check('zero cap',solve(rows,0),[])
check('empty rows',solve([],N),[])
check('cap beyond size',solve(rows,9),[[a,'x','y'],[b,'u','v'],[c,'m','n']])
check('distinct rejected alternatives',solve([[a,'x','y'],[b,'x','z']],3),[[a,'x','y'],[b,'x','z']])
check('one row',solve([[a,'x','y']],1),[[a,'x','y']])
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
dedup before cap	`[['a1', 'x', 'y']]`	`[['a1', 'x', 'y'], ['b1', 'u', 'v']]`	Failed
zero cap	`[]`	`[]`	Passed
empty rows	`[]`	`[]`	Passed
cap beyond size	`[['a1', 'x', 'y'], ['b1', 'u', 'v'], ['c1', 'm', 'n']]`	`[['a1', 'x', 'y'], ['b1', 'u', 'v'], ['c1', 'm', 'n']]`	Passed
distinct rejected alternatives	`[['a1', 'x', 'y']]`	`[['a1', 'x', 'y'], ['b1', 'x', 'z']]`	Failed
one row	`[['a1', 'x', 'y']]`	`[['a1', 'x', 'y']]`	Passed

SHA-256 / 94f7fe5031f92bc42a0a1487f8ac67f06996380f3e69debaf0513df4fc5c7b1e

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 2026-09-30T16:48:31.716477+00:00.

Case digest / 6cb43c8e8c33e5738187948bd4c23bac41bdf088e14bd66e3cdb6a6d91702060