Aller au contenu

Fangcong Yin

@fcyin.bsky.social did:plc:s43wvxidgqjtipupeiec4tjg

CS PhD student @UT Austin studying NLP. Prev:@CornellCIS

Sur Bluesky depuis le

Abonnés
153
Abonnements
316
Posts
2
Jours dans les classements
0

Résumé

Fangcong Yin (@fcyin.bsky.social) est sur Bluesky depuis le 19 novembre 2024. Le compte a 153 abonnés, suit 316 comptes et a publié 2 posts.

Présence dans les classements

Les meilleurs jours où les posts de ce compte sont entrés dans le top quotidien de Debya.

Ce compte n’est encore apparu dans aucun classement quotidien.

Posts aimés récemment

J’aime publics du réseau Bluesky.

  • Leqi Liu@leqiliu.bsky.social

    This has huge practical implications! It opens the door to using small, efficient models as sandboxes to probe, understand, and even steer their much larger counterparts. Paper: arxiv.org/abs/2506.00653 Joint work with Femi Bello, @anubrata.bsky.social, Fanzhi Zeng, @fcyin.bsky.social

    Voir le post
  • Juan Diego Rodriguez@juand-r.bsky.social

    Anyone know of any datasets of fair play detective fiction (i.e., no essential information is withheld from the reader), with annotations for (ideally) who the culprit(s) are, who the suspects are, and other relevant information (clues, red herrings, important reveals, narrative turning points)?

    Voir le post
  • Leqi Liu@leqiliu.bsky.social

    What if you could understand and control an LLM by studying its *smaller* sibling? Our new paper introduces the Linear Representation Transferability Hypothesis. We find that the internal representations of different-sized models can be translated into one another using a simple linear(affine) map.

    Voir le post
  • Ramya Namuduri@ramyanamuduri.bsky.social

    Have that eerie feeling of déjà vu when reading model-generated text 👀, but can’t pinpoint the specific words or phrases 👀? ✨We introduce QUDsim, to quantify discourse similarities beyond lexical, syntactic, and content overlap.

    Voir le post
  • Juan Diego Rodriguez@juand-r.bsky.social

    2.) [ICLR 2025] When does CoT help? It turns out that gains are mainly on math and symbolic reasoning. Check out our paper for a deep dive into MMLU, hundreds of experiments, and a meta-analysis of CoT across 3 conferences covering over 100 papers! arxiv.org/abs/2409.12183

    Voir le post

Connectez-vous pour en voir plus

Comptes bloqués

Sur Bluesky, les blocages sont publics.

Ce compte ne bloque personne.