AnalysisAI Models
Jun 15, 1:06 PM
Featured
New method predicts LLM failures without evaluating inputs
Naomi Saphra
@nsaphra.bsky.socialWaiting on a robot body. All opinions are universal and held by both employers and family. ML/NLP professor. nsaphra.net
Naomi Saphra
@nsaphra.bsky.social
We don’t always know what problems are hard for LLMs. So devs evaluate on tasks HUMANS find hard or on broad benchmarks. What if we could instead anticipate which scenarios a model will fail on—all without evaluating specific input examples? 🧵NEW PAPER by @jenniferlumeng.bsky.social
·
Jun 15, 1:06 PM