FA-99946 / Evaluation metric aggregation / Open access

Confusion counts omit false negatives · case 01

Confusion counts omit false negatives.

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

ROOT CAUSE

Only predicted-positive rows are accumulated, so missed positives vanish.

THE FAILURE

Only predicted-positive rows are accumulated, so missed positives vanish.

Unsuccessful approach: Deriving false negatives as total minus true positives counts true negatives too.

Case contract

Return [TP,FP,FN,TN] for binary labels 0 and 1.

Why this case matters

A small explicit evaluation contract makes aggregation behavior reproducible across labeled examples.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(truth, pred):
    tp=sum(a==1 and b==1 for a,b in zip(truth,pred)); fp=sum(a==0 and b==1 for a,b in zip(truth,pred)); tn=sum(a==0 and b==0 for a,b in zip(truth,pred)); return [tp,fp,0,tn]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('each cell',solve([1,1,0,0],[1,0,1,0]),[1,1,1,1])
check('all positive',solve([1]*N,[1]*N),[N,0,0,0])
check('all negative',solve([0]*N,[0]*N),[0,0,0,N])
check('empty',solve([],[]),[0,0,0,0])
check('missed positive',solve([1],[0]),[0,0,1,0])
check('false alarm',solve([0],[1]),[0,1,0,0])
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
each cell	`[1, 1, 0, 1]`	`[1, 1, 1, 1]`	Failed
all positive	`[1, 0, 0, 0]`	`[1, 0, 0, 0]`	Passed
all negative	`[0, 0, 0, 1]`	`[0, 0, 0, 1]`	Passed
empty	`[0, 0, 0, 0]`	`[0, 0, 0, 0]`	Passed
missed positive	`[0, 0, 0, 0]`	`[0, 0, 1, 0]`	Failed
false alarm	`[0, 1, 0, 0]`	`[0, 1, 0, 0]`	Passed

SHA-256 / 462965cc548cecb777ba41de240d2514fed4b0fda914987ab695059fcce194ce

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(truth, pred):
    tp=sum(a==1 and b==1 for a,b in zip(truth,pred)); fp=sum(a==0 and b==1 for a,b in zip(truth,pred)); tn=sum(a==0 and b==0 for a,b in zip(truth,pred)); return [tp,fp,len(truth)-tp,tn]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('each cell',solve([1,1,0,0],[1,0,1,0]),[1,1,1,1])
check('all positive',solve([1]*N,[1]*N),[N,0,0,0])
check('all negative',solve([0]*N,[0]*N),[0,0,0,N])
check('empty',solve([],[]),[0,0,0,0])
check('missed positive',solve([1],[0]),[0,0,1,0])
check('false alarm',solve([0],[1]),[0,1,0,0])
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
each cell	`[1, 1, 3, 1]`	`[1, 1, 1, 1]`	Failed
all positive	`[1, 0, 0, 0]`	`[1, 0, 0, 0]`	Passed
all negative	`[0, 0, 1, 1]`	`[0, 0, 0, 1]`	Failed
empty	`[0, 0, 0, 0]`	`[0, 0, 0, 0]`	Passed
missed positive	`[0, 0, 1, 0]`	`[0, 0, 1, 0]`	Passed
false alarm	`[0, 1, 1, 0]`	`[0, 1, 0, 0]`	Failed

SHA-256 / a6b2082cf4429d1e6697235bfaaed472a2e8939d3be605015740c099a3130418

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 finite toy labels and scores; this defines no benchmark protocol, production metric library, or model quality 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 2026-09-30T16:48:29.887669+00:00.

Case digest / 681eb7db0a22105e7749e239a4613bdcc30bae79250fab5063d13e04c00a0fed