Back to papers
October 7, 2026cs.LGcs.CLcs.CR

PatchBench: Measuring Collateral Damage in Activation Patching

Alexi Canesse, Mathis Le Bail, Maël Jenny, Clément Elliker, Mahammed El Sharkawy, Sonia Vanier
Read on arXivDownload PDF
Categories

cs.LG, cs.CL, cs.CR

PatchBench: Measuring Collateral Damage in Activation Patching - AI Research | One9Founders