What if you could control how long a reasoning model “thinks”? Presenting L1-1.5B, an RL-trained reasoning model with: - controllable thinking length via a prompt - better performance per token than S1 - better short CoT performance than GPT-4o cmu-l3.github.io/l1 🧵

Training with more data = better LLMs, right? 🚨 False! Scaling language models by adding more pre-training data can decrease your performance after post-training! Introducing "catastrophic overtraining." 🥁🧵+arXiv 👇 1/9

Excited to share our new paper, "One-Minute Video Generation with Test-Time Training (TTT)" in collaboration with NVIDIA. We augment a pre-trained Transformer with TTT-layers and finetune it to generate one-minute Tom and Jerry cartoons with strong temporal and spatial

🚨New preprint🚨 I’m super excited to share our work: An Empirical Study on Strong-Weak Model Collaboration for Repo-level Code Generation 📜: arxiv.org/abs/2505.20182 w/ Atharva Naik , Yiqing Xie and Carolyn Rose 🧵

Future AI systems interacting with humans will need to perform social reasoning that is grounded in behavioral cues and external knowledge. We introduce Social Genome to study and advance this form of reasoning in models! New paper w/ Marian Qian, Paul Liang, & LP Morency!

RepoST was accepted to Conference on Language Modeling !!! See you in Montreal 🚀 #COLM2025

Very excited to share that HAICosystem has been accepted to #COLM2025 ! 🎉 Multi-turn, interactive evaluation is THE future, think Tau-Bench, TheAgentCompany, Sotopia, ... Proud to take a small step toward open-ended, interactive AI safety eval, and excited for what’s next! 😎

Congrats, Shubham!! Well deserved!!!

Puzzlehunt lover🙋

At this morning's #colm2025 poster session, come see Yiqing Xie present a method for scalable construction of coding environments from GitHub repos. Poster 76 Paper: arxiv.org/abs/2503.07358 Work with Alex Xie, Divyanshu Sheth, Pengfei Liu, and Carolyn Rose

We just built and released the largest dataset for supervised fine-tuning of agentic LMs, 1.27M trajectories (~36B tokens)! Up until now, large-scale SFT for agents is rare - not for lack of data, but because of fragmentation across heterogeneous formats, tools, and interfaces.