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 ↗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
Quantization: Output zero point is scaled with the accumulator · case 01
The output zero point changes with the requantization multiplier.
Quantization: Output zero point is scaled with the accumulator · case 02
The output zero point changes with the requantization multiplier.
Quantization: Output zero point is scaled with the accumulator · case 03
The output zero point changes with the requantization multiplier.
Quantization: Output zero point is scaled with the accumulator · case 04
The output zero point changes with the requantization multiplier.
Quantization: Output zero point is scaled with the accumulator · case 05
The output zero point changes with the requantization multiplier.
Quantization: Bias is divided by the input scale · case 01
An output-unit bias changes when the input scale changes.
Quantization: Bias is divided by the input scale · case 02
An output-unit bias changes when the input scale changes.
Quantization: Bias is divided by the input scale · case 03
An output-unit bias changes when the input scale changes.
Quantization: Bias is divided by the input scale · case 04
An output-unit bias changes when the input scale changes.
Quantization: Bias is divided by the input scale · case 05
An output-unit bias changes when the input scale changes.
Quantization: Accumulator silently wraps at a narrow width · case 01
Accumulator silently wraps at a narrow width.
Quantization: Accumulator silently wraps at a narrow width · case 02
Accumulator silently wraps at a narrow width.
Quantization: Accumulator silently wraps at a narrow width · case 03
Accumulator silently wraps at a narrow width.
Quantization: Accumulator silently wraps at a narrow width · case 04
Accumulator silently wraps at a narrow width.
Quantization: Accumulator silently wraps at a narrow width · case 05
Accumulator silently wraps at a narrow width.
Quantization: Dot product omits input zero-point correction · case 01
Dot product omits input zero-point correction.
Quantization: Dot product omits input zero-point correction · case 02
Dot product omits input zero-point correction.
Quantization: Dot product omits input zero-point correction · case 03
Dot product omits input zero-point correction.
Quantization: Dot product omits input zero-point correction · case 04
Dot product omits input zero-point correction.
Quantization: Dot product omits input zero-point correction · case 05
Dot product omits input zero-point correction.
Quantization: Dot product silently drops unmatched tail · case 01
Dot product silently drops unmatched tail.
Quantization: Dot product silently drops unmatched tail · case 02
Dot product silently drops unmatched tail.
Quantization: Dot product silently drops unmatched tail · case 03
Dot product silently drops unmatched tail.
Quantization: Dot product silently drops unmatched tail · case 04
Dot product silently drops unmatched tail.
Quantization: Dot product silently drops unmatched tail · case 05
Dot product silently drops unmatched tail.
Quantization: Symmetric quantization retains nonzero zero point · case 01
Symmetric quantization retains nonzero zero point.
Quantization: Symmetric quantization retains nonzero zero point · case 02
Symmetric quantization retains nonzero zero point.
Quantization: Symmetric quantization retains nonzero zero point · case 03
Symmetric quantization retains nonzero zero point.
Quantization: Symmetric quantization retains nonzero zero point · case 04
Symmetric quantization retains nonzero zero point.
Quantization: Symmetric quantization retains nonzero zero point · case 05
Symmetric quantization retains nonzero zero point.
Quantization: Symmetric minimum code is asymmetric · case 01
Symmetric minimum code is asymmetric.
Quantization: Symmetric minimum code is asymmetric · case 02
Symmetric minimum code is asymmetric.
Quantization: Symmetric minimum code is asymmetric · case 03
Symmetric minimum code is asymmetric.
Quantization: Symmetric minimum code is asymmetric · case 04
Symmetric minimum code is asymmetric.
Quantization: Symmetric minimum code is asymmetric · case 05
Symmetric minimum code is asymmetric.
Quantization: Tensor scale is selected from the first value · case 01
Tensor scale is selected from the first value.
Quantization: Tensor scale is selected from the first value · case 02
Tensor scale is selected from the first value.
Quantization: Tensor scale is selected from the first value · case 03
Tensor scale is selected from the first value.
Quantization: Tensor scale is selected from the first value · case 04
Tensor scale is selected from the first value.
Quantization: Tensor scale is selected from the first value · case 05
Tensor scale is selected from the first value.
Tensor views: row-major stride recurrence · case 01
row-major stride recurrence.
Tensor views: row-major stride recurrence · case 02
row-major stride recurrence.
Tensor views: row-major stride recurrence · case 03
row-major stride recurrence.
Tensor views: row-major stride recurrence · case 04
row-major stride recurrence.
Tensor views: row-major stride recurrence · case 05
row-major stride recurrence.
Tensor views: multidimensional storage address · case 01
multidimensional storage address.
Tensor views: multidimensional storage address · case 02
multidimensional storage address.
Tensor views: multidimensional storage address · case 03
multidimensional storage address.
Tensor views: multidimensional storage address · case 04
multidimensional storage address.
Tensor views: multidimensional storage address · case 05
multidimensional storage address.
Tensor views: transpose swaps matching shape and stride axes · case 01
transpose swaps matching shape and stride axes.
Tensor views: transpose swaps matching shape and stride axes · case 02
transpose swaps matching shape and stride axes.
Tensor views: transpose swaps matching shape and stride axes · case 03
transpose swaps matching shape and stride axes.
Tensor views: transpose swaps matching shape and stride axes · case 04
transpose swaps matching shape and stride axes.
Tensor views: transpose swaps matching shape and stride axes · case 05
transpose swaps matching shape and stride axes.
Tensor views: slice origin advances storage offset · case 01
slice origin advances storage offset.
Tensor views: slice origin advances storage offset · case 02
slice origin advances storage offset.
Tensor views: slice origin advances storage offset · case 03
slice origin advances storage offset.
Tensor views: slice origin advances storage offset · case 04
slice origin advances storage offset.
Tensor views: slice origin advances storage offset · case 05
slice origin advances storage offset.
Tensor views: broadcasted singleton gets zero stride · case 01
broadcasted singleton gets zero stride.
Tensor views: broadcasted singleton gets zero stride · case 02
broadcasted singleton gets zero stride.
Tensor views: broadcasted singleton gets zero stride · case 03
broadcasted singleton gets zero stride.
Tensor views: broadcasted singleton gets zero stride · case 04
broadcasted singleton gets zero stride.
Tensor views: broadcasted singleton gets zero stride · case 05
broadcasted singleton gets zero stride.
Tensor views: reshape inferred dimension uses exact element count · case 01
reshape inferred dimension uses exact element count.
Tensor views: reshape inferred dimension uses exact element count · case 02
reshape inferred dimension uses exact element count.
Tensor views: reshape inferred dimension uses exact element count · case 03
reshape inferred dimension uses exact element count.
Tensor views: reshape inferred dimension uses exact element count · case 04
reshape inferred dimension uses exact element count.
Tensor views: reshape inferred dimension uses exact element count · case 05
reshape inferred dimension uses exact element count.
Tensor views: reshape preserves total element count · case 01
reshape preserves total element count.
Tensor views: reshape preserves total element count · case 02
reshape preserves total element count.
Tensor views: reshape preserves total element count · case 03
reshape preserves total element count.
Tensor views: reshape preserves total element count · case 04
reshape preserves total element count.
Tensor views: reshape preserves total element count · case 05
reshape preserves total element count.
Tensor views: reshape view requires coalescible strides · case 01
reshape view requires coalescible strides.
Tensor views: reshape view requires coalescible strides · case 02
reshape view requires coalescible strides.
Tensor views: reshape view requires coalescible strides · case 03
reshape view requires coalescible strides.
Tensor views: reshape view requires coalescible strides · case 04
reshape view requires coalescible strides.
Tensor views: reshape view requires coalescible strides · case 05
reshape view requires coalescible strides.
Tensor views: flatten view combines adjacent strides · case 01
flatten view combines adjacent strides.
Tensor views: flatten view combines adjacent strides · case 02
flatten view combines adjacent strides.
Tensor views: flatten view combines adjacent strides · case 03
flatten view combines adjacent strides.
Tensor views: flatten view combines adjacent strides · case 04
flatten view combines adjacent strides.
Tensor views: flatten view combines adjacent strides · case 05
flatten view combines adjacent strides.
Tensor views: axis permutation preserves stride pairing · case 01
axis permutation preserves stride pairing.
Tensor views: axis permutation preserves stride pairing · case 02
axis permutation preserves stride pairing.
Tensor views: axis permutation preserves stride pairing · case 03
axis permutation preserves stride pairing.
Tensor views: axis permutation preserves stride pairing · case 04
axis permutation preserves stride pairing.
Tensor views: axis permutation preserves stride pairing · case 05
axis permutation preserves stride pairing.
Tensor views: unsqueeze inserts a unit dimension at normalized axis · case 01
unsqueeze inserts a unit dimension at normalized axis.
Tensor views: unsqueeze inserts a unit dimension at normalized axis · case 02
unsqueeze inserts a unit dimension at normalized axis.
Tensor views: unsqueeze inserts a unit dimension at normalized axis · case 03
unsqueeze inserts a unit dimension at normalized axis.
Tensor views: unsqueeze inserts a unit dimension at normalized axis · case 04
unsqueeze inserts a unit dimension at normalized axis.
Tensor views: unsqueeze inserts a unit dimension at normalized axis · case 05
unsqueeze inserts a unit dimension at normalized axis.
Tensor views: squeeze removes only unit dimensions · case 01
squeeze removes only unit dimensions.
Tensor views: squeeze removes only unit dimensions · case 02
squeeze removes only unit dimensions.
Tensor views: squeeze removes only unit dimensions · case 03
squeeze removes only unit dimensions.
Tensor views: squeeze removes only unit dimensions · case 04
squeeze removes only unit dimensions.
Tensor views: squeeze removes only unit dimensions · case 05
squeeze removes only unit dimensions.
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 ↗