New! 📜🚨
Mamba, Griffin, RWKV, RetNet, Recurrent Gemma- 2024 is the year of gated linear RNNs! What's their secret sauce? Is it necessary to develop unique interpretability tools for each layer separately? Our paper provides a unified framework to answer these questions! [1/5]

📣📣 Super proud to present the most exciting project of my PhD so far: “HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models”.

HippoRAG, as the title suggests, is a brain-inspired RAG framework that enables LLMs to effectively and efficiently

Super excited to introduce HippoRAG, a method I enjoyed developing the most in 2024. It’s led by my amazing student Bernal Bernal Jiménez and joint with Yiheng Shu Yu Gu Michi Yasunaga. Bernal’s thread gives a good technical account, so I’ll just share some personal thoughts

iVideoGPT: Interactive VideoGPTs are Scalable World Models

Achieves competitive perf with SotA method on action-conditioned video prediction, visual planning, and model-based RL

arxiv.org/abs/2405.15223

thanks AK for sharing our work led by the amazing Boshi Wang from OSU NLP Group. This is one of my favorite work so far. I'm excited that several important topics: 1) inductive learning of latent deduction rules, 2) implicit/parametric reasoning of neural networks, and 3)

Happy Friday!

The next big leap in language modeling is going to come from finetuning them on agent trajectories. This will lead to a big accuracy improvement in end-to-end programming (e.g. SWE-agent), controlling desktop apps (e.g. OSWorld) and web browsing (e.g. SeeAct).

[Sorry, there was a mistake on the project name in our new large multimodal model post.😓 We did reformat  the multimodal project as XGen-MM as part of our XGen initiative. We are now reposting the project with the updated name. Thanks.]

We introduce XGen-MM, a series of large

I am thrilled to announce that I will be joining Duke University
Duke CompSci as an Assistant Professor in summer 2025. Super excited for the next chapter! Stay tuned for the launch of my lab 🧠🤖

@Unitree H1 @ICLR2024 very robust humanoid with strong policy

Exciting efficient method on fill-in-the-middle codegen models!

The SWE-agent preprint has finally landed! Check it out at swe-agent.com/paper.pdf

I'll be at #ICLR2024 🇦🇹 to support our works:
1. Text2Reward: Reward Shaping with Language Models for Reinforcement Learning iclr.cc/virtual/2024/p…
2. Lemur: Harmonizing Natural Language and Code for Language Agents iclr.cc/virtual/2024/p…

Excited to meet old and new friends!

Really exciting to see how far it can go in this way!!!

Introducing DrEureka🎓, our latest effort pushing the frontier of robot learning using LLMs!

DrEureka uses LLMs to automatically design reward functions and tune physics parameters to enable sim-to-real robot learning. DrEureka can propose effective sim-to-real configurations

I will present my thesis defense tomorrow!

Language Agents: From Next-Token Prediction to Digital Automation

- 10am EST on Thursday, May 2
- princeton.zoom.us/my/shunyuy
- WebShop, ReAct, ToT, CoALA
- Briefly: SWE-bench/agent
- Thoughts on the future of language agents

🙌🙌🙌