🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Interacting with the external world and reacting based on outcomes are crucial capabilities of agentic systems, but existing LLMs’ ability to do so is limited. Introducing Paprika 🌶️, our work on making LLMs general decision makers than can solve new tasks zero-shot. 🧵 1/n

1/Being in academia is such a privilege: You get to collaborate with insanely talented & passionate students on their journey to upskill themselves. Very excited to share *OpenUnlearning*: a unified, easily extensible framework for unlearning led by Anmol Mekala Vineeth🧵

Are current reasoning models optimal for test-time scaling? 🌠 No! Models make the same incorrect guess over and over again. We show that you can fix this problem w/o any crazy tricks 💫 – just do weight ensembling (WiSE-FT) for big gains on math! 1/N

Huge thank you to NVIDIA Data Center for gifting a brand new #NVIDIADGX B200 to CMU’s Catalyst Research Group! This AI supercomputing system will afford Catalyst the ability to run and test their work on a world-class unified AI platform.

Thanks NVIDIA Data Center for the DGX B200 machine for the CMU Catalyst group! I'm perhaps already a bit too enthralled by it in the photos...

spicy Zico Kolter

When we train models to do QA, are we robustly improving context dependency? No! In our ICLR Oral (Fri 11 AM), we show that if the base model knows the facts already, it shortcuts and learns to ignore the context completely! Visit us to learn more about knowledge conflicts 😀

✨ Love 4o-style image generation but prefer to use Midjourney? Tired of manual prompt crafting from inspo images? PRISM to the rescue! 🖼️→📝→🖼️ We automate black-box prompt engineering—no training, no embeddings, just accurate, readable prompts from your inspo images! 1/🧵

Looking forward to giving a talk this Friday OpenAI with Zhili Feng on some of our privacy & memorization research + how it applies to production LLMs! We've been gaining momentum on detecting, quantifying & erasing memorization; excited to explore its real-world impact!

A shorter version of the first three chapters of my thesis is accepted by ICML 2025. It provides a quick start for those interested in learning about the contexture theory. Check it out: arxiv.org/abs/2505.01557

Excited to be talking today about how research into memorization provides a fundamentally different lens on safety!

Excited to share our work with my amazing collaborators, Goodeat, Xingjian Bai, Zico Kolter, and Kaiming. In a word, we show an “identity learning” approach for generative modeling, by relating the instantaneous/average velocity in an identity. The resulting model,

✨ Did you know that NOT using all generated rollouts in GRPO can boost your reasoning LLM? Meet PODS! We down-sample rollouts and train on just a fraction, delivering notable gains over vanilla GRPO. (1/7)

Wrote my first blog post! I wanted to share a powerful yet under-recognized way to develop emotional maturity as a researcher: making it a habit to read about the ✨past ✨ and learn from it to make sense of the present

🚨Excited to release OS-Harm! 🚨 The safety of computer use agents has been largely overlooked. We created a new safety benchmark based on OSWorld for measuring 3 broad categories of harm: 1. deliberate user misuse, 2. prompt injections, 3. model misbehavior.

now the code is up here: github.com/Gsunshine/mean…

A mental model I find useful: all data acquisition (web scrapes, synthetic data, RL rollouts, etc.) is really an exploration problem 🔍. This perspective has some interesting implications for where AI is heading. Wrote down some thoughts: yidingjiang.github.io/blog/post/expl…