Mor Geva (@megamor2) 's Twitter Profile
Mor Geva

@megamor2

ID: 850356925535531009

linkhttps://mega002.github.io/ calendar_today07-04-2017 14:37:44

368 Tweet

1,1K Followers

474 Following

Mor Geva (@megamor2) 's Twitter Profile Photo
Mor Geva
@megamor2

5 months ago

Do you have a "tell" when you are about to lie? We find that LLMs have “tells” in their internal representations which allow estimating how knowledgeable a model is about an entity 𝘣𝘦𝘧𝘰𝘳𝘦 it generates even a single token. Paper: arxiv.org/abs/2406.12673… 🧵 Daniela Gottesman

Do you have a "tell" when you are about to lie? We find that LLMs have “tells” in their internal representations which allow estimating how knowledgeable a model is about an entity 𝘣𝘦𝘧𝘰𝘳𝘦 it generates even a single token. Paper: arxiv.org/abs/2406.12673… 🧵 <a href="/dhgottesman/">Daniela Gottesman</a>