🔥 "Vibe coding" is everywhere—but is it really care-free? We introduce 𝐑𝐞𝐚𝐋, an RL framework that trains LLMs with automated program analysis feedback, enabling "vibe coding" to be not just fast—but 𝐯𝐮𝐥𝐧𝐞𝐫𝐚𝐛𝐢𝐥𝐢𝐭𝐲-𝐟𝐫𝐞𝐞 & 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧-𝐫𝐞𝐚𝐝𝐲 🛡️

Using retrieval? --> check out this work by my awesome collaborator on how to increase diversity when retrieving!

🚨 Code is live! Check out LoRe – a modular, lightweight codebase for personalized reward modeling from user preferences. 📦 Few-shot personalization 📊 Benchmarks: TLDR, PRISM, PersonalLLM 👉 github.com/facebookresear… Huge thanks to AI at Meta for open-sourcing this research 🙌

😵‍💫 Struggling with 𝐟𝐢𝐧𝐞-𝐭𝐮𝐧𝐢𝐧𝐠 𝐌𝐨𝐄? Meet 𝐃𝐞𝐧𝐬𝐞𝐌𝐢𝐱𝐞𝐫 — an MoE post-training method that offers more 𝐩𝐫𝐞𝐜𝐢𝐬𝐞 𝐫𝐨𝐮𝐭𝐞𝐫 𝐠𝐫𝐚𝐝𝐢𝐞𝐧𝐭, making MoE 𝐞𝐚𝐬𝐢𝐞𝐫 𝐭𝐨 𝐭𝐫𝐚𝐢𝐧 and 𝐛𝐞𝐭𝐭𝐞𝐫 𝐩𝐞𝐫𝐟𝐨𝐫𝐦𝐢𝐧𝐠! Blog: fengyao.notion.site/moe-posttraini…

Are you compositionally curious 🤓 Want to know how to learn embeddings using🌲? In our new #ICML2025 paper, we present Banyan: A recursive net that you can train super efficiently for any language or domain, and get embeddings competitive with much much larger LLMs 1/🧵

I will be at the Actionable Interpretability Workshop (Actionable Interpretability Workshop ICML2025, #ICML) presenting *SSAEs* in the East Ballroom A from 1-2pm. Drop by (or send a DM) to chat about (actionable) interpretability, (actionable) identifiability, and everything in between!

Transformers struggle with length generalization and long context. What can we do about it? Our new #TMLR paper with Roland Fernandez , Paul Smolensky and Jianfeng Gao shows how to handle the issue. Using a new attention mechanism called TRA. Curious? Read the 🧵 for more 🤓

Llama Nemotron model just got Super-Charged ⚡️We released Llama-Nemotron-Super-v1.5 today! The best open model that can be deployed on a single H100 🚀 Enhanced for reasoning, tool use, general chat, and instruction following. HF : huggingface.co/nvidia/Llama-3…

Glad to share that our paper was accepted the main EMNLP 2025 Conference! x.com/Sahil1V/status…

RFDiffusion3 generates all atom bound conformation, making it significant for flexible targets like DNA. An excellent teamwork to achieve something impossible by any one of us in just few months. Jasper Butcher Rohith Krishna biorxiv.org/content/10.110…

Very excited to share our paper "De novo Design of All-atom Biomolecular Interactions with RFdiffusion3", now on BioRXiv. biorxiv.org/content/10.110… 1/n

Hands down Hila is one of the best advisors out there, you would be lucky to work with her! (as was I :) )

[1/9] While pretraining data might be hitting a wall, novel methods for modeling it are just getting started! We introduce future summary prediction (FSP), where the model predicts future sequence embeddings to reduce teacher forcing & shortcut learning. 📌Predict a learned

Sarah is great, go work with her!

Today, we report a method for design of active enzymes, RFdiffusion2, in Nature Methods. For the first time, we are able to design enzymes with native-range catalytic activity. We also are releasing our next frontier model, RFdiffusion3, code 👇