1/3 Today, an anecdote shared by an invited speaker at #NeurIPS2024 left many Chinese scholars, myself included, feeling uncomfortable. As a community, I believe we should take a moment to reflect on why such remarks in public discourse can be offensive and harmful.

[Long Tweet Ahead] Faculty Interview Tips & Common Questions: 🧘‍♀️0. Firstly, do not be nervous - Almost everything can be prepared in advance:) - Be grateful for everyone's time. - Think of it as an opportunity to share your research with others -- exciting, right? - Technical

🚀 Some great news to you: Reward is boosting! Reward curve is AI's ultimate language!!!

🚀 Introducing RAGEN—the world’s first reproduction of DeepSeek-R1(-Zero) methods for training agentic AI models! We’re betting big on the future of RL + LLM + Agents 🤖✨. This release is a minimally viable leap toward that vision. Code and more intro 🔗:

I'm so excited that my lab mates were able to produce such groundbreaking work in such a short time. A big salute to Zihan Wang - on RAGEN 🫡, who literally didn't sleep during the weekend.

Surprise finding: Our simplified AICO version actually outperforms TRICO on Sokoban, likely because of the game's sparse rewards only for a successful run🤯 But TRICO shows superior exploration skills, cracking the toughest puzzles while AICO stably handles simpler ones 🧩 (7/n)

Super excited to introduce VAGEN!! We trained a 3B VLM agent in Sokoban and it can sometimes solve 6-step game! Honored be part of the team!

No visual models one can survive the challenge, but... ... ... ... ... Our VAGEN can we are doing small progress but visual agent has yet even more to do

And our tiny VAGEN:

We are embarrassed to say that VAGEN is the No. 1 visual agent framework, but... it's true X Post: x.com/wzihanw/status… Blog: mll-lab.notion.site/vagen Code: github.com/RAGEN-AI/VAGEN

🚀 Introducing T* and LV-Haystack — our latest leap forward in VLMs for long video understanding! 🧩 Lightweight plugin: T* boosting LLaVA-OV-72B (56→62%) and GPT-4o (50→53%)! ⚡ Fast inference: 34.9s → 10.4s latency, 691 → 170 TFLOPs v.s. SOTA. 📚 Large-scale dataset: 400

Introducing T* and LV-Haystack -- targeting needle-in-the-haystack for long videos! 🤗 LV-Haystack annotated 400+ hours of videos and 15,000+ samples. 🧩 Lightweight plugin for any proprietary and open-source VLMs: T* boosting LLaVA-OV-72B [56→62%] and GPT-4o [50→53%] within

Why does your RL training always collapse? In our new paper of RAGEN, we explore what breaks when you train LLM *Agents* with multi-turn reinforcement learning—and possibly how to fix it. 📄 github.com/RAGEN-AI/RAGEN… 🌐 ragen-ai.github.io 1/🧵👇

We are very excited announcing our MLL lab! We are looking for collaborators on RAGEN, VAGEN, Chain-of-experts, T*, LongVideoHaystack, foundation models for embodied agents, etc mll-lab-nu.github.io

Finally available on arxiv! arxiv.org/abs/2506.18945

Can VLMs build Spatial Mental Models like humans? Reasoning from limited views? Reasoning from partial observations? Reasoning about unseen objects behind furniture / beyond current view? Check out MindCube! 🌐mll-lab-nu.github.io/mind-cube/ 📰arxiv.org/pdf/2506.21458