FAILURE MAP

Understand the failure.
Verify the repair.

Small, reproducible software failures. The broken implementation, the fix that didn’t work, and the one that passed—preserved together.

Explore the cases ↓How results are verified ↗
100840Executable case variants
20168Distinct failure mechanisms
302520Executed implementations
20168Open-access cases

WHAT THE ARCHIVE CONTAINS

100840 executable cases. 20168 are open.

Every case records the implementation that fails, the fix that did not work, and the repair that passed its checks—with recorded outputs and source hashes. This release adds 100840 cases across 20168 failure mechanisms and 254 domains.

The open tier gives you the failure and the unsuccessful fix for one case in every mechanism. The remaining 80672 cases, 5 variants per mechanism, are member-only: the verified repair, its recorded checks, and the full fixture suite are held in the member archive. Read the methodology ↗

A RECORD OF WHAT WENT WRONG

Browse the archive / 100840

Python · Standard library
REFERENCEFAILURE MECHANISMDOMAINACCESS
FA-17001

Exponential divided difference reverses its exponent separation · case 01

Exponential divided difference reverses its exponent separation.

Floating-point arithmetic● Open access↗
FA-17002

Exponential divided difference reverses its exponent separation · case 02

Exponential divided difference reverses its exponent separation.

Floating-point arithmetic◈ Members↗
FA-17003

Exponential divided difference reverses its exponent separation · case 03

Exponential divided difference reverses its exponent separation.

Floating-point arithmetic◈ Members↗
FA-17004

Exponential divided difference reverses its exponent separation · case 04

Exponential divided difference reverses its exponent separation.

Floating-point arithmetic◈ Members↗
FA-17005

Exponential divided difference reverses its exponent separation · case 05

Exponential divided difference reverses its exponent separation.

Floating-point arithmetic◈ Members↗
FA-17006

Log divided difference subtracts logarithms of adjacent large values · case 01

Log divided difference subtracts logarithms of adjacent large values.

Floating-point arithmetic● Open access↗
FA-17007

Log divided difference subtracts logarithms of adjacent large values · case 02

Log divided difference subtracts logarithms of adjacent large values.

Floating-point arithmetic◈ Members↗
FA-17008

Log divided difference subtracts logarithms of adjacent large values · case 03

Log divided difference subtracts logarithms of adjacent large values.

Floating-point arithmetic◈ Members↗
FA-17009

Log divided difference subtracts logarithms of adjacent large values · case 04

Log divided difference subtracts logarithms of adjacent large values.

Floating-point arithmetic◈ Members↗
FA-17010

Log divided difference subtracts logarithms of adjacent large values · case 05

Log divided difference subtracts logarithms of adjacent large values.

Floating-point arithmetic◈ Members↗
FA-17011

Log divided difference uses the logarithm as its coincident limit · case 01

Log divided difference uses the logarithm as its coincident limit.

Floating-point arithmetic● Open access↗
FA-17012

Log divided difference uses the logarithm as its coincident limit · case 02

Log divided difference uses the logarithm as its coincident limit.

Floating-point arithmetic◈ Members↗
FA-17013

Log divided difference uses the logarithm as its coincident limit · case 03

Log divided difference uses the logarithm as its coincident limit.

Floating-point arithmetic◈ Members↗
FA-17014

Log divided difference uses the logarithm as its coincident limit · case 04

Log divided difference uses the logarithm as its coincident limit.

Floating-point arithmetic◈ Members↗
FA-17015

Log divided difference uses the logarithm as its coincident limit · case 05

Log divided difference uses the logarithm as its coincident limit.

Floating-point arithmetic◈ Members↗
FA-17016

Log divided difference normalizes its increment by the destination · case 01

Log divided difference normalizes its increment by the destination.

Floating-point arithmetic● Open access↗
FA-17017

Log divided difference normalizes its increment by the destination · case 02

Log divided difference normalizes its increment by the destination.

Floating-point arithmetic◈ Members↗
FA-17018

Log divided difference normalizes its increment by the destination · case 03

Log divided difference normalizes its increment by the destination.

Floating-point arithmetic◈ Members↗
FA-17019

Log divided difference normalizes its increment by the destination · case 04

Log divided difference normalizes its increment by the destination.

Floating-point arithmetic◈ Members↗
FA-17020

Log divided difference normalizes its increment by the destination · case 05

Log divided difference normalizes its increment by the destination.

Floating-point arithmetic◈ Members↗
FA-17021

Log divided difference uses log1p even when the normalized decrement rounds to minus one · case 01

Log divided difference uses log1p even when the normalized decrement rounds to minus one.

Floating-point arithmetic● Open access↗
FA-17022

Log divided difference uses log1p even when the normalized decrement rounds to minus one · case 02

Log divided difference uses log1p even when the normalized decrement rounds to minus one.

Floating-point arithmetic◈ Members↗
FA-17023

Log divided difference uses log1p even when the normalized decrement rounds to minus one · case 03

Log divided difference uses log1p even when the normalized decrement rounds to minus one.

Floating-point arithmetic◈ Members↗
FA-17024

Log divided difference uses log1p even when the normalized decrement rounds to minus one · case 04

Log divided difference uses log1p even when the normalized decrement rounds to minus one.

Floating-point arithmetic◈ Members↗
FA-17025

Log divided difference uses log1p even when the normalized decrement rounds to minus one · case 05

Log divided difference uses log1p even when the normalized decrement rounds to minus one.

Floating-point arithmetic◈ Members↗
FA-17026

Log divided difference drops the sign of its denominator · case 01

Log divided difference drops the sign of its denominator.

Floating-point arithmetic● Open access↗
FA-17027

Log divided difference drops the sign of its denominator · case 02

Log divided difference drops the sign of its denominator.

Floating-point arithmetic◈ Members↗
FA-17028

Log divided difference drops the sign of its denominator · case 03

Log divided difference drops the sign of its denominator.

Floating-point arithmetic◈ Members↗
FA-17029

Log divided difference drops the sign of its denominator · case 04

Log divided difference drops the sign of its denominator.

Floating-point arithmetic◈ Members↗
FA-17030

Log divided difference drops the sign of its denominator · case 05

Log divided difference drops the sign of its denominator.

Floating-point arithmetic◈ Members↗
FA-17031

Log ratio materializes an overflowing or underflowing quotient · case 01

Log ratio materializes an overflowing or underflowing quotient.

Floating-point arithmetic● Open access↗
FA-17032

Log ratio materializes an overflowing or underflowing quotient · case 02

Log ratio materializes an overflowing or underflowing quotient.

Floating-point arithmetic◈ Members↗
FA-17033

Log ratio materializes an overflowing or underflowing quotient · case 03

Log ratio materializes an overflowing or underflowing quotient.

Floating-point arithmetic◈ Members↗
FA-17034

Log ratio materializes an overflowing or underflowing quotient · case 04

Log ratio materializes an overflowing or underflowing quotient.

Floating-point arithmetic◈ Members↗
FA-17035

Log ratio materializes an overflowing or underflowing quotient · case 05

Log ratio materializes an overflowing or underflowing quotient.

Floating-point arithmetic◈ Members↗
FA-17036

Log ratio subtracts close logarithms at large magnitude · case 01

Log ratio subtracts close logarithms at large magnitude.

Floating-point arithmetic● Open access↗
FA-17037

Log ratio subtracts close logarithms at large magnitude · case 02

Log ratio subtracts close logarithms at large magnitude.

Floating-point arithmetic◈ Members↗
FA-17038

Log ratio subtracts close logarithms at large magnitude · case 03

Log ratio subtracts close logarithms at large magnitude.

Floating-point arithmetic◈ Members↗
FA-17039

Log ratio subtracts close logarithms at large magnitude · case 04

Log ratio subtracts close logarithms at large magnitude.

Floating-point arithmetic◈ Members↗
FA-17040

Log ratio subtracts close logarithms at large magnitude · case 05

Log ratio subtracts close logarithms at large magnitude.

Floating-point arithmetic◈ Members↗
FA-17041

Log ratio normalizes its difference by the numerator · case 01

Log ratio normalizes its difference by the numerator.

Floating-point arithmetic● Open access↗
FA-17042

Log ratio normalizes its difference by the numerator · case 02

Log ratio normalizes its difference by the numerator.

Floating-point arithmetic◈ Members↗
FA-17043

Log ratio normalizes its difference by the numerator · case 03

Log ratio normalizes its difference by the numerator.

Floating-point arithmetic◈ Members↗
FA-17044

Log ratio normalizes its difference by the numerator · case 04

Log ratio normalizes its difference by the numerator.

Floating-point arithmetic◈ Members↗
FA-17045

Log ratio normalizes its difference by the numerator · case 05

Log ratio normalizes its difference by the numerator.

Floating-point arithmetic◈ Members↗
FA-17046

Log ratio reverses the distant logarithm difference · case 01

Log ratio reverses the distant logarithm difference.

Floating-point arithmetic● Open access↗
FA-17047

Log ratio reverses the distant logarithm difference · case 02

Log ratio reverses the distant logarithm difference.

Floating-point arithmetic◈ Members↗
FA-17048

Log ratio reverses the distant logarithm difference · case 03

Log ratio reverses the distant logarithm difference.

Floating-point arithmetic◈ Members↗
FA-17049

Log ratio reverses the distant logarithm difference · case 04

Log ratio reverses the distant logarithm difference.

Floating-point arithmetic◈ Members↗
FA-17050

Log ratio reverses the distant logarithm difference · case 05

Log ratio reverses the distant logarithm difference.

Floating-point arithmetic◈ Members↗
FA-17051

Log ratio accepts zero endpoints as ordinary positives · case 01

Log ratio accepts zero endpoints as ordinary positives.

Floating-point arithmetic● Open access↗
FA-17052

Log ratio accepts zero endpoints as ordinary positives · case 02

Log ratio accepts zero endpoints as ordinary positives.

Floating-point arithmetic◈ Members↗
FA-17053

Log ratio accepts zero endpoints as ordinary positives · case 03

Log ratio accepts zero endpoints as ordinary positives.

Floating-point arithmetic◈ Members↗
FA-17054

Log ratio accepts zero endpoints as ordinary positives · case 04

Log ratio accepts zero endpoints as ordinary positives.

Floating-point arithmetic◈ Members↗
FA-17055

Log ratio accepts zero endpoints as ordinary positives · case 05

Log ratio accepts zero endpoints as ordinary positives.

Floating-point arithmetic◈ Members↗
FA-17056

Closeness rejects equal infinities before checking exact equality · case 01

Closeness rejects equal infinities before checking exact equality.

Floating-point arithmetic● Open access↗
FA-17057

Closeness rejects equal infinities before checking exact equality · case 02

Closeness rejects equal infinities before checking exact equality.

Floating-point arithmetic◈ Members↗
FA-17058

Closeness rejects equal infinities before checking exact equality · case 03

Closeness rejects equal infinities before checking exact equality.

Floating-point arithmetic◈ Members↗
FA-17059

Closeness rejects equal infinities before checking exact equality · case 04

Closeness rejects equal infinities before checking exact equality.

Floating-point arithmetic◈ Members↗
FA-17060

Closeness rejects equal infinities before checking exact equality · case 05

Closeness rejects equal infinities before checking exact equality.

Floating-point arithmetic◈ Members↗
FA-17061

Closeness accepts an infinite distance against infinite tolerance · case 01

Closeness accepts an infinite distance against infinite tolerance.

Floating-point arithmetic● Open access↗
FA-17062

Closeness accepts an infinite distance against infinite tolerance · case 02

Closeness accepts an infinite distance against infinite tolerance.

Floating-point arithmetic◈ Members↗
FA-17063

Closeness accepts an infinite distance against infinite tolerance · case 03

Closeness accepts an infinite distance against infinite tolerance.

Floating-point arithmetic◈ Members↗
FA-17064

Closeness accepts an infinite distance against infinite tolerance · case 04

Closeness accepts an infinite distance against infinite tolerance.

Floating-point arithmetic◈ Members↗
FA-17065

Closeness accepts an infinite distance against infinite tolerance · case 05

Closeness accepts an infinite distance against infinite tolerance.

Floating-point arithmetic◈ Members↗
FA-17066

Relative closeness scales by only the first operand · case 01

Relative closeness scales by only the first operand.

Floating-point arithmetic● Open access↗
FA-17067

Relative closeness scales by only the first operand · case 02

Relative closeness scales by only the first operand.

Floating-point arithmetic◈ Members↗
FA-17068

Relative closeness scales by only the first operand · case 03

Relative closeness scales by only the first operand.

Floating-point arithmetic◈ Members↗
FA-17069

Relative closeness scales by only the first operand · case 04

Relative closeness scales by only the first operand.

Floating-point arithmetic◈ Members↗
FA-17070

Relative closeness scales by only the first operand · case 05

Relative closeness scales by only the first operand.

Floating-point arithmetic◈ Members↗
FA-17071

Relative closeness uses signed magnitudes · case 01

Relative closeness uses signed magnitudes.

Floating-point arithmetic● Open access↗
FA-17072

Relative closeness uses signed magnitudes · case 02

Relative closeness uses signed magnitudes.

Floating-point arithmetic◈ Members↗
FA-17073

Relative closeness uses signed magnitudes · case 03

Relative closeness uses signed magnitudes.

Floating-point arithmetic◈ Members↗
FA-17074

Relative closeness uses signed magnitudes · case 04

Relative closeness uses signed magnitudes.

Floating-point arithmetic◈ Members↗
FA-17075

Relative closeness uses signed magnitudes · case 05

Relative closeness uses signed magnitudes.

Floating-point arithmetic◈ Members↗
FA-17076

Closeness sums tolerances instead of choosing the larger · case 01

Closeness sums tolerances instead of choosing the larger.

Floating-point arithmetic● Open access↗
FA-17077

Closeness sums tolerances instead of choosing the larger · case 02

Closeness sums tolerances instead of choosing the larger.

Floating-point arithmetic◈ Members↗
FA-17078

Closeness sums tolerances instead of choosing the larger · case 03

Closeness sums tolerances instead of choosing the larger.

Floating-point arithmetic◈ Members↗
FA-17079

Closeness sums tolerances instead of choosing the larger · case 04

Closeness sums tolerances instead of choosing the larger.

Floating-point arithmetic◈ Members↗
FA-17080

Closeness sums tolerances instead of choosing the larger · case 05

Closeness sums tolerances instead of choosing the larger.

Floating-point arithmetic◈ Members↗
FA-17081

Closeness excludes values exactly at tolerance · case 01

Closeness excludes values exactly at tolerance.

Floating-point arithmetic● Open access↗
FA-17082

Closeness excludes values exactly at tolerance · case 02

Closeness excludes values exactly at tolerance.

Floating-point arithmetic◈ Members↗
FA-17083

Closeness excludes values exactly at tolerance · case 03

Closeness excludes values exactly at tolerance.

Floating-point arithmetic◈ Members↗
FA-17084

Closeness excludes values exactly at tolerance · case 04

Closeness excludes values exactly at tolerance.

Floating-point arithmetic◈ Members↗
FA-17085

Closeness excludes values exactly at tolerance · case 05

Closeness excludes values exactly at tolerance.

Floating-point arithmetic◈ Members↗
FA-17086

Closeness silently takes absolute values of negative tolerances · case 01

Closeness silently takes absolute values of negative tolerances.

Floating-point arithmetic● Open access↗
FA-17087

Closeness silently takes absolute values of negative tolerances · case 02

Closeness silently takes absolute values of negative tolerances.

Floating-point arithmetic◈ Members↗
FA-17088

Closeness silently takes absolute values of negative tolerances · case 03

Closeness silently takes absolute values of negative tolerances.

Floating-point arithmetic◈ Members↗
FA-17089

Closeness silently takes absolute values of negative tolerances · case 04

Closeness silently takes absolute values of negative tolerances.

Floating-point arithmetic◈ Members↗
FA-17090

Closeness silently takes absolute values of negative tolerances · case 05

Closeness silently takes absolute values of negative tolerances.

Floating-point arithmetic◈ Members↗
FA-17091

Opposite-sign midpoint overflows its endpoint difference · case 01

Opposite-sign midpoint overflows its endpoint difference.

Floating-point arithmetic● Open access↗
FA-17092

Opposite-sign midpoint overflows its endpoint difference · case 02

Opposite-sign midpoint overflows its endpoint difference.

Floating-point arithmetic◈ Members↗
FA-17093

Opposite-sign midpoint overflows its endpoint difference · case 03

Opposite-sign midpoint overflows its endpoint difference.

Floating-point arithmetic◈ Members↗
FA-17094

Opposite-sign midpoint overflows its endpoint difference · case 04

Opposite-sign midpoint overflows its endpoint difference.

Floating-point arithmetic◈ Members↗
FA-17095

Opposite-sign midpoint overflows its endpoint difference · case 05

Opposite-sign midpoint overflows its endpoint difference.

Floating-point arithmetic◈ Members↗
FA-17096

Same-sign midpoint overflows its endpoint sum · case 01

Same-sign midpoint overflows its endpoint sum.

Floating-point arithmetic● Open access↗
FA-17097

Same-sign midpoint overflows its endpoint sum · case 02

Same-sign midpoint overflows its endpoint sum.

Floating-point arithmetic◈ Members↗
FA-17098

Same-sign midpoint overflows its endpoint sum · case 03

Same-sign midpoint overflows its endpoint sum.

Floating-point arithmetic◈ Members↗
FA-17099

Same-sign midpoint overflows its endpoint sum · case 04

Same-sign midpoint overflows its endpoint sum.

Floating-point arithmetic◈ Members↗
FA-17100

Same-sign midpoint overflows its endpoint sum · case 05

Same-sign midpoint overflows its endpoint sum.

Floating-point arithmetic◈ Members↗

INSPECTABLE BY DESIGN

Every result has a runnable source.

Runnable implementations with recorded outputs, source hashes, and explicit contracts. Related variants share a failure mechanism and belong together in evaluation splits.

Read the methodology ↗