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-15901

Leaving zero increments its old signed encoding · case 01

Leaving zero increments its old signed encoding.

Floating-point arithmetic● Open access↗
FA-15902

Leaving zero increments its old signed encoding · case 02

Leaving zero increments its old signed encoding.

Floating-point arithmetic◈ Members↗
FA-15903

Leaving zero increments its old signed encoding · case 03

Leaving zero increments its old signed encoding.

Floating-point arithmetic◈ Members↗
FA-15904

Leaving zero increments its old signed encoding · case 04

Leaving zero increments its old signed encoding.

Floating-point arithmetic◈ Members↗
FA-15905

Leaving zero increments its old signed encoding · case 05

Leaving zero increments its old signed encoding.

Floating-point arithmetic◈ Members↗
FA-15906

Positive adjacency always increments the encoding · case 01

Positive adjacency always increments the encoding.

Floating-point arithmetic● Open access↗
FA-15907

Positive adjacency always increments the encoding · case 02

Positive adjacency always increments the encoding.

Floating-point arithmetic◈ Members↗
FA-15908

Positive adjacency always increments the encoding · case 03

Positive adjacency always increments the encoding.

Floating-point arithmetic◈ Members↗
FA-15909

Positive adjacency always increments the encoding · case 04

Positive adjacency always increments the encoding.

Floating-point arithmetic◈ Members↗
FA-15910

Positive adjacency always increments the encoding · case 05

Positive adjacency always increments the encoding.

Floating-point arithmetic◈ Members↗
FA-15911

Negative adjacency uses unsigned monotonic order · case 01

Negative adjacency uses unsigned monotonic order.

Floating-point arithmetic● Open access↗
FA-15912

Negative adjacency uses unsigned monotonic order · case 02

Negative adjacency uses unsigned monotonic order.

Floating-point arithmetic◈ Members↗
FA-15913

Negative adjacency uses unsigned monotonic order · case 03

Negative adjacency uses unsigned monotonic order.

Floating-point arithmetic◈ Members↗
FA-15914

Negative adjacency uses unsigned monotonic order · case 04

Negative adjacency uses unsigned monotonic order.

Floating-point arithmetic◈ Members↗
FA-15915

Negative adjacency uses unsigned monotonic order · case 05

Negative adjacency uses unsigned monotonic order.

Floating-point arithmetic◈ Members↗
FA-15916

ULP lookup reports NaN as infinity · case 01

ULP lookup reports NaN as infinity.

Floating-point arithmetic● Open access↗
FA-15917

ULP lookup reports NaN as infinity · case 02

ULP lookup reports NaN as infinity.

Floating-point arithmetic◈ Members↗
FA-15918

ULP lookup reports NaN as infinity · case 03

ULP lookup reports NaN as infinity.

Floating-point arithmetic◈ Members↗
FA-15919

ULP lookup reports NaN as infinity · case 04

ULP lookup reports NaN as infinity.

Floating-point arithmetic◈ Members↗
FA-15920

ULP lookup reports NaN as infinity · case 05

ULP lookup reports NaN as infinity.

Floating-point arithmetic◈ Members↗
FA-15921

ULP lookup assigns a finite spacing to infinity · case 01

ULP lookup assigns a finite spacing to infinity.

Floating-point arithmetic● Open access↗
FA-15922

ULP lookup assigns a finite spacing to infinity · case 02

ULP lookup assigns a finite spacing to infinity.

Floating-point arithmetic◈ Members↗
FA-15923

ULP lookup assigns a finite spacing to infinity · case 03

ULP lookup assigns a finite spacing to infinity.

Floating-point arithmetic◈ Members↗
FA-15924

ULP lookup assigns a finite spacing to infinity · case 04

ULP lookup assigns a finite spacing to infinity.

Floating-point arithmetic◈ Members↗
FA-15925

ULP lookup assigns a finite spacing to infinity · case 05

ULP lookup assigns a finite spacing to infinity.

Floating-point arithmetic◈ Members↗
FA-15926

Zero ULP is rounded to zero · case 01

Zero ULP is rounded to zero.

Floating-point arithmetic● Open access↗
FA-15927

Zero ULP is rounded to zero · case 02

Zero ULP is rounded to zero.

Floating-point arithmetic◈ Members↗
FA-15928

Zero ULP is rounded to zero · case 03

Zero ULP is rounded to zero.

Floating-point arithmetic◈ Members↗
FA-15929

Zero ULP is rounded to zero · case 04

Zero ULP is rounded to zero.

Floating-point arithmetic◈ Members↗
FA-15930

Zero ULP is rounded to zero · case 05

Zero ULP is rounded to zero.

Floating-point arithmetic◈ Members↗
FA-15931

ULP uses the exponent of the next binade · case 01

ULP uses the exponent of the next binade.

Floating-point arithmetic● Open access↗
FA-15932

ULP uses the exponent of the next binade · case 02

ULP uses the exponent of the next binade.

Floating-point arithmetic◈ Members↗
FA-15933

ULP uses the exponent of the next binade · case 03

ULP uses the exponent of the next binade.

Floating-point arithmetic◈ Members↗
FA-15934

ULP uses the exponent of the next binade · case 04

ULP uses the exponent of the next binade.

Floating-point arithmetic◈ Members↗
FA-15935

ULP uses the exponent of the next binade · case 05

ULP uses the exponent of the next binade.

Floating-point arithmetic◈ Members↗
FA-15936

ULP scaling underflows below the subnormal spacing · case 01

ULP scaling underflows below the subnormal spacing.

Floating-point arithmetic● Open access↗
FA-15937

ULP scaling underflows below the subnormal spacing · case 02

ULP scaling underflows below the subnormal spacing.

Floating-point arithmetic◈ Members↗
FA-15938

ULP scaling underflows below the subnormal spacing · case 03

ULP scaling underflows below the subnormal spacing.

Floating-point arithmetic◈ Members↗
FA-15939

ULP scaling underflows below the subnormal spacing · case 04

ULP scaling underflows below the subnormal spacing.

Floating-point arithmetic◈ Members↗
FA-15940

ULP scaling underflows below the subnormal spacing · case 05

ULP scaling underflows below the subnormal spacing.

Floating-point arithmetic◈ Members↗
FA-15941

An inexact result ignores sticky-only residue · case 01

An inexact result ignores sticky-only residue.

Floating-point arithmetic● Open access↗
FA-15942

An inexact result ignores sticky-only residue · case 02

An inexact result ignores sticky-only residue.

Floating-point arithmetic◈ Members↗
FA-15943

An inexact result ignores sticky-only residue · case 03

An inexact result ignores sticky-only residue.

Floating-point arithmetic◈ Members↗
FA-15944

An inexact result ignores sticky-only residue · case 04

An inexact result ignores sticky-only residue.

Floating-point arithmetic◈ Members↗
FA-15945

An inexact result ignores sticky-only residue · case 05

An inexact result ignores sticky-only residue.

Floating-point arithmetic◈ Members↗
FA-15946

Nearest-even rounding ignores retained parity · case 01

Nearest-even rounding ignores retained parity.

Floating-point arithmetic● Open access↗
FA-15947

Nearest-even rounding ignores retained parity · case 02

Nearest-even rounding ignores retained parity.

Floating-point arithmetic◈ Members↗
FA-15948

Nearest-even rounding ignores retained parity · case 03

Nearest-even rounding ignores retained parity.

Floating-point arithmetic◈ Members↗
FA-15949

Nearest-even rounding ignores retained parity · case 04

Nearest-even rounding ignores retained parity.

Floating-point arithmetic◈ Members↗
FA-15950

Nearest-even rounding ignores retained parity · case 05

Nearest-even rounding ignores retained parity.

Floating-point arithmetic◈ Members↗
FA-15951

Nearest rounding treats any discarded bit as above-half · case 01

Nearest rounding treats any discarded bit as above-half.

Floating-point arithmetic● Open access↗
FA-15952

Nearest rounding treats any discarded bit as above-half · case 02

Nearest rounding treats any discarded bit as above-half.

Floating-point arithmetic◈ Members↗
FA-15953

Nearest rounding treats any discarded bit as above-half · case 03

Nearest rounding treats any discarded bit as above-half.

Floating-point arithmetic◈ Members↗
FA-15954

Nearest rounding treats any discarded bit as above-half · case 04

Nearest rounding treats any discarded bit as above-half.

Floating-point arithmetic◈ Members↗
FA-15955

Nearest rounding treats any discarded bit as above-half · case 05

Nearest rounding treats any discarded bit as above-half.

Floating-point arithmetic◈ Members↗
FA-15956

Nearest rounding loses the round bit after a guard · case 01

Nearest rounding loses the round bit after a guard.

Floating-point arithmetic● Open access↗
FA-15957

Nearest rounding loses the round bit after a guard · case 02

Nearest rounding loses the round bit after a guard.

Floating-point arithmetic◈ Members↗
FA-15958

Nearest rounding loses the round bit after a guard · case 03

Nearest rounding loses the round bit after a guard.

Floating-point arithmetic◈ Members↗
FA-15959

Nearest rounding loses the round bit after a guard · case 04

Nearest rounding loses the round bit after a guard.

Floating-point arithmetic◈ Members↗
FA-15960

Nearest rounding loses the round bit after a guard · case 05

Nearest rounding loses the round bit after a guard.

Floating-point arithmetic◈ Members↗
FA-15961

Nearest rounding loses sticky residue after a guard · case 01

Nearest rounding loses sticky residue after a guard.

Floating-point arithmetic● Open access↗
FA-15962

Nearest rounding loses sticky residue after a guard · case 02

Nearest rounding loses sticky residue after a guard.

Floating-point arithmetic◈ Members↗
FA-15963

Nearest rounding loses sticky residue after a guard · case 03

Nearest rounding loses sticky residue after a guard.

Floating-point arithmetic◈ Members↗
FA-15964

Nearest rounding loses sticky residue after a guard · case 04

Nearest rounding loses sticky residue after a guard.

Floating-point arithmetic◈ Members↗
FA-15965

Nearest rounding loses sticky residue after a guard · case 05

Nearest rounding loses sticky residue after a guard.

Floating-point arithmetic◈ Members↗
FA-15966

Nearest-away incorrectly follows ties-to-even · case 01

Nearest-away incorrectly follows ties-to-even.

Floating-point arithmetic● Open access↗
FA-15967

Nearest-away incorrectly follows ties-to-even · case 02

Nearest-away incorrectly follows ties-to-even.

Floating-point arithmetic◈ Members↗
FA-15968

Nearest-away incorrectly follows ties-to-even · case 03

Nearest-away incorrectly follows ties-to-even.

Floating-point arithmetic◈ Members↗
FA-15969

Nearest-away incorrectly follows ties-to-even · case 04

Nearest-away incorrectly follows ties-to-even.

Floating-point arithmetic◈ Members↗
FA-15970

Nearest-away incorrectly follows ties-to-even · case 05

Nearest-away incorrectly follows ties-to-even.

Floating-point arithmetic◈ Members↗
FA-15971

Rounding upward increments negative magnitudes · case 01

Rounding upward increments negative magnitudes.

Floating-point arithmetic● Open access↗
FA-15972

Rounding upward increments negative magnitudes · case 02

Rounding upward increments negative magnitudes.

Floating-point arithmetic◈ Members↗
FA-15973

Rounding upward increments negative magnitudes · case 03

Rounding upward increments negative magnitudes.

Floating-point arithmetic◈ Members↗
FA-15974

Rounding upward increments negative magnitudes · case 04

Rounding upward increments negative magnitudes.

Floating-point arithmetic◈ Members↗
FA-15975

Rounding upward increments negative magnitudes · case 05

Rounding upward increments negative magnitudes.

Floating-point arithmetic◈ Members↗
FA-15976

Rounding downward increments positive magnitudes · case 01

Rounding downward increments positive magnitudes.

Floating-point arithmetic● Open access↗
FA-15977

Rounding downward increments positive magnitudes · case 02

Rounding downward increments positive magnitudes.

Floating-point arithmetic◈ Members↗
FA-15978

Rounding downward increments positive magnitudes · case 03

Rounding downward increments positive magnitudes.

Floating-point arithmetic◈ Members↗
FA-15979

Rounding downward increments positive magnitudes · case 04

Rounding downward increments positive magnitudes.

Floating-point arithmetic◈ Members↗
FA-15980

Rounding downward increments positive magnitudes · case 05

Rounding downward increments positive magnitudes.

Floating-point arithmetic◈ Members↗
FA-15981

Truncation applies the nearest rounding increment · case 01

Truncation applies the nearest rounding increment.

Floating-point arithmetic● Open access↗
FA-15982

Truncation applies the nearest rounding increment · case 02

Truncation applies the nearest rounding increment.

Floating-point arithmetic◈ Members↗
FA-15983

Truncation applies the nearest rounding increment · case 03

Truncation applies the nearest rounding increment.

Floating-point arithmetic◈ Members↗
FA-15984

Truncation applies the nearest rounding increment · case 04

Truncation applies the nearest rounding increment.

Floating-point arithmetic◈ Members↗
FA-15985

Truncation applies the nearest rounding increment · case 05

Truncation applies the nearest rounding increment.

Floating-point arithmetic◈ Members↗
FA-15986

Rounding increments after applying the negative sign · case 01

Rounding increments after applying the negative sign.

Floating-point arithmetic● Open access↗
FA-15987

Rounding increments after applying the negative sign · case 02

Rounding increments after applying the negative sign.

Floating-point arithmetic◈ Members↗
FA-15988

Rounding increments after applying the negative sign · case 03

Rounding increments after applying the negative sign.

Floating-point arithmetic◈ Members↗
FA-15989

Rounding increments after applying the negative sign · case 04

Rounding increments after applying the negative sign.

Floating-point arithmetic◈ Members↗
FA-15990

Rounding increments after applying the negative sign · case 05

Rounding increments after applying the negative sign.

Floating-point arithmetic◈ Members↗
FA-15991

Encoding flushes signed zero to positive zero · case 01

Encoding flushes signed zero to positive zero.

Floating-point arithmetic● Open access↗
FA-15992

Encoding flushes signed zero to positive zero · case 02

Encoding flushes signed zero to positive zero.

Floating-point arithmetic◈ Members↗
FA-15993

Encoding flushes signed zero to positive zero · case 03

Encoding flushes signed zero to positive zero.

Floating-point arithmetic◈ Members↗
FA-15994

Encoding flushes signed zero to positive zero · case 04

Encoding flushes signed zero to positive zero.

Floating-point arithmetic◈ Members↗
FA-15995

Encoding flushes signed zero to positive zero · case 05

Encoding flushes signed zero to positive zero.

Floating-point arithmetic◈ Members↗
FA-15996

Overflow uses the largest finite value as its threshold · case 01

Overflow uses the largest finite value as its threshold.

Floating-point arithmetic● Open access↗
FA-15997

Overflow uses the largest finite value as its threshold · case 02

Overflow uses the largest finite value as its threshold.

Floating-point arithmetic◈ Members↗
FA-15998

Overflow uses the largest finite value as its threshold · case 03

Overflow uses the largest finite value as its threshold.

Floating-point arithmetic◈ Members↗
FA-15999

Overflow uses the largest finite value as its threshold · case 04

Overflow uses the largest finite value as its threshold.

Floating-point arithmetic◈ Members↗
FA-16000

Overflow uses the largest finite value as its threshold · case 05

Overflow uses the largest finite value as its threshold.

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 ↗