ICLR 2025

Our work maps hereditary relationships between models. We find that weights🏋️ are sufficient for decoding the Origin of Models🌳

Presenting today at @iclr_conf, 15:00–17:30, Hall 3, poster #360.

Come by to see our method, visualizations, and interactive demo atlas👀🚀 #ICLR2025 pic.twitter.com/PooRwBb16t

— Eliahu Horwitz (@EliahuHorwitz) April 25, 2025

#ICLR2025 Oral

How can we control LMs using diverse signals such as static analyses, test cases, and simulations?
In our paper “Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo” we:
Cast controlled generation as an inference problem, with the LM… pic.twitter.com/wUe3KduBwg

— João Loula (@JoaoLoula) April 25, 2025

ICLR Spotlighted! ✨@canfer_akbulut presenting our research tomorrow at #ICLR2025 on evaluating #AI's understanding of historical images using our new Century dataset. pic.twitter.com/YiEslLifsd

— Verena Rieser (@verena_rieser) April 25, 2025

🚨Announcing our #ICLR2025 Oral!

🔥Diffusion LMs are on the rise for parallel text generation! But unlike autoregressive LMs, they struggle with quality, fixed-length constraints & lack of KV caching.

🚀Introducing Block Diffusion—combining autoregressive and diffusion models… pic.twitter.com/cY7BbmWJwE

— Marianne Arriola @ ICLR’25 (@mariannearr) March 13, 2025