WhatWorkedBench

What is WhatWorkedBench?

Evaluates AI agents on predicting component-change effects in 36 executable workflow tasks across 8 families using 30 data sources and 1,248 native outcomes.

Released

2026-09-22

Evaluates

Language & Knowledge, Software & AI Compute, cs.AI

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://github.com/EthanNing/WhatWorkedBench

WhatWorkedBench code

• Code https://github.com/EthanNing/WhatWorkedBench
• Website https://ethanning.github.io/WhatWorkedBench/

Which sources cite WhatWorkedBench?

• WhatWorkedBench

Related Language & Knowledge, Software & AI Compute, cs.AI benchmarks

• 2d-fet-bench
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• AARRI-Bench
• ABACUS
• ABC-Bench
• Active-SWE
• ActiveSciBench