Yi Wu
@jxwuyi
AI/RL researcher, Assistant Prof. at @Tsinghua_Uni, leading the RL lab at @AntResearch_, PhD at @berkeley_ai, frequent flyer and milk tea lover.
We release fully async RL system AReaL-boba² for LLM & SOTA code RL w. Qwen3-14B! @Alibaba_Qwen #opensource 🚀system&algorithm co-design → 2.77x faster ✅ 69.1 on LiveCodeBench 🔥 multi-turn RL ready 🔗 Project: github.com/inclusionAI/AR… 📄 Paper: arxiv.org/pdf/2505.24298 1/3👇
Come to Ant Group/Research Booth at ICLR2025 for some of the most exciting T-shirts ever in (my) AI conference history! Grab our AReaL-boba T-shirt! Ping me if you want a chat (or boba!) We are hiring! @InclusionAI666 #ICLR2025 #AGI
This work is done by my PhD student Zijian and researchers at Nvidia! @LigengZhu @songhan_mit Big congrats! We are also integrating VLM into AReaL.
🚀Summer Fest Day 4: Turbocharging Vision-Language Models with SGLang + NVILA 4.4× throughput, 2.2× faster response time! We've integrated NVILA into SGLang, enabling high-performance, scalable serving of vision-language models. This unlocks a 4.4× TPS boost and significantly…
A great VLM with fairly small size. It can be a new base model for further RL training to boost its agentic capacities.
We @Zai_org are thrilled to open-source GLM-4.1V-9B-Thinking, a VLM that can think with long CoTs. SoTA in <10B VLMs, comparable to Qwen-2.5-VL-72B in 18 tasks. One RL to rule them all! Details - Tech report: arxiv.org/abs/2507.01006 - Code: github.com/THUDM/GLM-4.1V…
Human & robot dog cooperative soccer game with multi-agent RL! Check out the video ;)
7️⃣/7️⃣ 📄 Paper: arxiv.org/abs/2505.13834 🎬 Full video: youtu.be/7gq7N16jKgI 🙏 Big thanks to my co-author Yuman Gao and to our mentors @ZhongyuLi4, @jxwuyi, and @KoushilSreenath for their invaluable guidance and support! #Robotics #Quadruped #RobotSoccer
My first 3 PhD students successfully make their defense! How time flies! #PhDone #Graduation2025
AReaL team at ICLR! Happy post conference holiday (and deadline)! #ICLR2025
Check out our AReaL talk at ICLR 2024! Apr 24, 1-2pm, Garnet Room #ICLR2025 Come to chat! Our team is Hiring!
1 day to go！See you guys in #ICLR2025 @jxwuyi @gujinjie
See you guys in Singapore! Let’s catch up and grab a milk tea ;)
We will attend the #ICRL 2025 in Singapore. Meet AReal and AWorld at the event,fully open-sourced projects for LLMs from RL reasoning to agent. @jxwuyi @BenGU206961 ⏰：24 April 🚩： Singapore Expo，Room Garnet 216-218 See you there! More about AReal and AWorld：…
We will attend the #ICRL 2025 in Singapore. Meet AReal and AWorld at the event,fully open-sourced projects for LLMs from RL reasoning to agent. @jxwuyi @BenGU206961 ⏰：24 April 🚩： Singapore Expo，Room Garnet 216-218 See you there! More about AReal and AWorld：…
Thank you so much for sharing @AdinaYakup! Two clarifications. We achieve sota math reasoning only on 7B models. For QwQ-32b, 200 samples leads to comparable numbers specifically on AIME2024. But our AReaL-boba system is indeed fast and stable! 🧋🧋
AReal-Boba 🔥 a fully open RL Frameworks released by AntGroup, an affiliate company of Alibaba. ✨ 7B/32B - Apache2.0 ✨ Outperform on math reasoning ✨ Replicating QwQ-32B with 200 data under $200 ✨ All-in-one: weights, datasets, code & tech report huggingface.co/collections/in…
AReaL v0.1.1 released! In addition to various system stability improvements, we also now have an extremely user friendly tutorial for 1-click running AReaL at public cloud nodes. Check out the tutorial: github.com/inclusionAI/AR…