Zihan Wang - on RAGEN @wzihanw
PhD Student @NorthwesternU. Intern @yutori_ai. I study PhysiCS of LLM. Ex @deepseek_ai @uiuc_nlp @RUC. RAGEN | Chain-of-Experts | ESFT. zihanwang314.github.io Joined March 2022-
Tweets865
-
Followers23K
-
Following609
-
Likes3K
Check out the 1st Behavior Challenge, co-host with our Foundation Models for Embodied Agent Challenge at NeurIPS …models-meet-embodied-agents.github.io/behavior_chall… When I first moved my focus from LLMs/VLMs toward embodied agents, I expected the biggest challenges would be around perception, motor…
Check out the 1st Behavior Challenge, co-host with our Foundation Models for Embodied Agent Challenge at NeurIPS …models-meet-embodied-agents.github.io/behavior_chall… When I first moved my focus from LLMs/VLMs toward embodied agents, I expected the biggest challenges would be around perception, motor…
For agents to improve over time, they can’t afford to forget what they’ve already mastered. We found that supervised fine-tuning forgets more than RL when training on a new task! Want to find out why? 👇
.@yutori_ai's Co-CEO @deviparikh on why she left @Meta to start her company with @abhshkdz. “We are rethinking what the interface with the web looks like, and that requires a certain sort of… thinking from first principles, thinking from scratch, thinking from a clean…
(1/N) How close are we to enabling robots to solve the long-horizon, complex tasks that matter in everyday life? 🚨 We are thrilled to invite you to join the 1st BEHAVIOR Challenge @NeurIPS 2025, submission deadline: 11/15. 🏆 Prizes: 🥇 $1,000 🥈 $500 🥉 $300
Since everyone is talking about RL Environments and GRPO now but no one knows how it works we thought it would be cool to make an explainer video + code you can run: This is an example of using GRPO to train Qwen 2.5 to play 2048 (code in thread) 🧵:
Introducing DeepSeek-V3.1: our first step toward the agent era! 🚀 🧠 Hybrid inference: Think & Non-Think — one model, two modes ⚡️ Faster thinking: DeepSeek-V3.1-Think reaches answers in less time vs. DeepSeek-R1-0528 🛠️ Stronger agent skills: Post-training boosts tool use and…
All week during rebuttals, I have started each day with the same reminder: stay humble, stay kind, don't let this turn me mean. When I was doing PhD, reviewers never felt this mean. There is a bright-eyed student sitting on the other side, and such reviews will destroy…
To guys diving into fine-tuning open-source MoEs today: check out ESFT, our customized PEFT method for MoE models. Train with 90% less parameters, gain 95%+ task perf and keep 98% general perf :)
To guys diving into fine-tuning open-source MoEs today: check out ESFT, our customized PEFT method for MoE models. Train with 90% less parameters, gain 95%+ task perf and keep 98% general perf :)
Excited to do a talk at Agentic AI Summit! Session 4: Foundations of Agents 📍 Frontier Stage 📅 2:45pm PT Will talk about "RAGEN: Training Agents via Reinforcing Reasoning", including: - How to monitor RL training to converge for multi-turn LLM Agents? RAGEN…
Excited to do a talk at Agentic AI Summit! Session 4: Foundations of Agents 📍 Frontier Stage 📅 2:45pm PT Will talk about "RAGEN: Training Agents via Reinforcing Reasoning", including: - How to monitor RL training to converge for multi-turn LLM Agents? RAGEN… https://t.co/L0QMLs4M1R
🏆Thrilled to receive @aclmeeting Inaugural Dissertation Award Honorable Mention. “Multimodality” has moved incredibly fast that my PhD research already feels like from a different era. It makes me wonder how challenging and anxious for today’s students to choose thesis…
🏆Thrilled to receive @aclmeeting Inaugural Dissertation Award Honorable Mention. “Multimodality” has moved incredibly fast that my PhD research already feels like from a different era. It makes me wonder how challenging and anxious for today’s students to choose thesis… https://t.co/8VxKWM90Lt
Congratulations to my amazing advisor @ManlingLi_ !!!
Congratulations to my amazing advisor @ManlingLi_ !!!
Is your LLM getting stuck while training with RL for agentic/ reasoning tasks? Well, turns out that a simple intuition of “trying again” works surprisingly well for reinforcement learning of LLMs and for domain adaptation!! Joint work with @NorthwesternEng @uwcse
Is your LLM getting stuck while training with RL for agentic/ reasoning tasks? Well, turns out that a simple intuition of “trying again” works surprisingly well for reinforcement learning of LLMs and for domain adaptation!! Joint work with @NorthwesternEng @uwcse
Do you find RL makes the LLM reasoning more stubborn? Keep repeating the same answers? How to make multi-turn conversational history be helpful in RL training? We identify a simple "try again" feedback can boost reasoning and make RL training a conversational manner!…
Do you find RL makes the LLM reasoning more stubborn? Keep repeating the same answers? How to make multi-turn conversational history be helpful in RL training? We identify a simple "try again" feedback can boost reasoning and make RL training a conversational manner!… https://t.co/fgnH0ffUcF
Single-turn data -> Multi-turn RL for general reasoning. Simple method yet effective. Worth trying out! Kudos to our amazing intern @liulicheng10 . He is applying for PhD positions for Fall 26!
Single-turn data -> Multi-turn RL for general reasoning. Simple method yet effective. Worth trying out! Kudos to our amazing intern @liulicheng10 . He is applying for PhD positions for Fall 26!
Excited that @RuohanZhang76 is joining NU @northwesterncs ! If you are thinking about pursuing a PhD, definitely reach out to him! During my wonderful year at @StanfordAILab @StanfordSVL, when I was completely new to robotics, he was the nicest person who was incredibly patient…
Excited that @RuohanZhang76 is joining NU @northwesterncs ! If you are thinking about pursuing a PhD, definitely reach out to him! During my wonderful year at @StanfordAILab @StanfordSVL, when I was completely new to robotics, he was the nicest person who was incredibly patient… https://t.co/iHgaF7HbSx
Excited to be speaking about agents at #GenAIWeek Silicon Valley this Sunday, July 13! Meet us at the panel Agent-to-Human Interfaces & Interactions, 10:50-11:30 AM :)
📣 Excited to announce SpaVLE: #NeurIPS2025 Workshop on Space in Vision, Language, and Embodied AI! 👉 …vision-language-embodied-ai.github.io 🦾Co-organized with an incredible team → @fredahshi · @maojiayuan · @DJiafei · @ManlingLi_ · David Hsu · @Kordjamshidi 🌌 Why Space & SpaVLE? We…
zeeshan saeed @zeeshansae62805
1 Followers 81 Following Tech entrepreneur & digital marketing strategist | Coffee addict ☕ | Building bridges between Islamabad & Silicon Valley | #Startups #AI #PakistanTech
Oqieke @Oqieke310
126 Followers 3K Following
Iploreerit @Iploreerit1945
6 Followers 200 Following
Shoaib Ansari @ShoaibA72236590
5 Followers 104 Following Multan businessman & agriculture tech enthusiast | Sustainable farming advocate | Sufi music lover | Connecting tradition with innovation | #AgTech #Multan #Sus
anbangliu @anbangliu
0 Followers 13 Following
Sabouhi @rjsabouhi
18 Followers 768 Following Hallucinations ≠ missing data. They’re drift through a weak constraint field. Δσ = γ(t) · ∇C(s, t) | Lock phase, raise gain, shape curvature.
Sebastian Sosa @Sebasti54919704
65 Followers 100 Following Exploring the extraordinary systems of the world.
QueStionEvrything @TimRussell
191 Followers 624 Following STICKS AND STONES BREAK BONES! NOT WORDS PEOPLE!!! Respect others freedom to pursue happiness! Keep an open mind and Question Everything!
2ro❤️☀️💚 @rooorro8
78 Followers 471 Following
geeked asl @smallprout_____
71 Followers 447 Following
enric @enric1909
141 Followers 623 Following Eu farei 10x se for preciso. Eles não estão preparados.
Lille @LilleB0t
227 Followers 513 Following Lille c'est bien. Bot à plein temps. Tous les #notfrombot sont tweetés par un humain. Fait avec amour pour sa région chez @BBotMaker.
diospremia @diospremia
175 Followers 823 Following
SEAI @SEAI_Agents
91 Followers 609 Following A unified platform to discover, catalog, monitor, and govern all your data—seamlessly, securely, and in one place.
SOSE @SOSE_Hq
85 Followers 571 Following Socket Security : #1 in software supply chain security. Next-gen SCA, SBOM, and zero-day threat prevention—devs love it.
FIight Rising @FLIRI_Games
347 Followers 666 Following Make friends, conquer monsters, and build the ultimate dragon clan! Play, grow, and rule the skies together. 🐉🔥
Lioden Game @Lioden_game
217 Followers 624 Following Lioden is a unique SIM game where you become the king of a lion pride—breed cubs, battle rivals, and rule the wild. Ready to lead?
Andrew I. Anderson @AndrewIAnderso1
74 Followers 438 Following
云创兽Ai @Doouipi8854154
1 Followers 21 Following 🚀 hot stock investing? That’s me, a smart girl! thrilled to connect. DM me for Nasdaq trends! 💡 #Gold #Investing
FEI LAN @FEILAN549587
40 Followers 654 Following
张洋 @touwenameng
46 Followers 3K Following
CryptoNerdCn 🦇🔊 @cryptonerdcn
13K Followers 1K Following InfoFi加分站: https://t.co/cOXs0xHtJY | Delegate of @ensdomains | 中文&日本語&English | @WasmCairo | Into @starknet by @starknetastrocn
jesus λnτ҈nι҉ @ThingJesus
27 Followers 144 Following
Serraw @Serraw046436
41 Followers 2K Following
Yusuf Duran Ocak @duran_ocakyusuf
3 Followers 201 Following I'm a humble, caring, honest and hardworking man... I work as a Banker with HalkBank of Turkey
Rui Yu @192802321y
2 Followers 7 Following
Mubashar @Mubasha197
2 Followers 19 Following
Uday @udaypandey
13 Followers 12 Following
Yuvraj Singh @YuvrajS9886
2K Followers 526 Following Ex - @turboml, @puch_ai | @iitmadras (left), @iiserkol, @UofMaryland, AIISC | YESIST '24 Finalist | Multimodal LLMs Research| Building SmolHub ☺️, NeatRL
Vadim Liventsev @vadimdotme
46 Followers 786 Following lead hip hop engineer @ https://t.co/y80r2iuMsN
zuko @KingZukox
1 Followers 436 Following
Allen @Allen94704
2 Followers 57 Following
getCream.AI @getcreamai
85 Followers 616 Following
Jasleen Kaur Dhanoa @jasleeen_dhanoa
5 Followers 196 Following
Summer Khan @summerkhan054
0 Followers 147 Following
Sebastian Sosa @Sebasti54919704
65 Followers 100 Following Exploring the extraordinary systems of the world.
parth @parth220_
1K Followers 1K Following Spiritual Ronin | agent whisperer @hud_evals, yc w25, human @earth | 👨🏽💻⚡️🌱
Crystal @crystalsssup
11K Followers 598 Following Staff @Kimi_Moonshot prev. co-maker of ModelizeAI & gemsouls "Personality goes a long way" @UCSanDiego
Denghui Zhang @denghui_zhang
476 Followers 674 Following Assistant Professor at Stevens Institute of Technology.
Avi Sil @aviaviavi__
2K Followers 533 Following Senior Director, Applied Science @Oracle | Past - Manager + Principal Scientist @IBMResearch AI | Tweets are my own opinion
Ruohan Zhang @RuohanZhang76
2K Followers 787 Following Postdoc @stanfordsvl @StanfordAILab; robot, brain, art; climbing, soccer, cooking, music, dance
Miki Pokryvailo @mpokryva
36 Followers 62 Following Eng at @yutori_ai. Previously founding eng @niledatabase.
Asher Dayanim @AshDayanim
49 Followers 115 Following Interested in the world, AI, Economics, and history Posts in English | עברית | فارسی | тоҷикӣ Opinions are my own
黄小茜Chelsey @_Chelsey404
5K Followers 728 Following 别走了，我知道你认识我 视频号 @黄小茜Chelsey Built https://t.co/6sbZbb7RHw | Quant Quant Quant | Columbia alum ♔
kristi edleson @kristiedleson
263 Followers 117 Following antisocial extrovert. will attend, won't initiate. 🪩 crossword enthusiast. building @yutori_ai
Xiaokang(Koe) Ye @koe_ye40329
10 Followers 63 Following 26' MSECE track in machine learning and data science @ UCSD
Krunal Shah @_krunalshah_
46 Followers 165 Following Building Agents @yutori_ai. Previously: @GoogleDeepMind @upennnlp @IITDelhi
Chenxin An @AnChancy46881
631 Followers 506 Following PhD Candidate @ HKUNLP Awardee of Hong Kong PhD Fellowship Scheme
Simon Yu @simon_ycl
519 Followers 785 Following 1st Year PhD Student, supervised by @shi_weiyan | Incoming intern in @OrbyAI | MRes and BSc Student @EdinburghNLP | Member of @CohereForAI
Nikolay Savinov 🇺�... @SavinovNikolay
3K Followers 0 Following Research Scientist at @GoogleDeepMind Work on LLM pre-training in Gemini ♊ Lead 10M context length in Gemini 1.5 📈
Rajko Radovanović @rajko_rad
6K Followers 5K Following AI/infra @a16z (partner to amazing teams eg @MistralAI @udiomusic @LumaLabsAI @cursor_ai @theworldlabs @braintrustdata) - alt @rajko_alt
Xiuyu Li @xiuyu_l
2K Followers 637 Following Efficiently scaling agents. CS PhD student @berkeley_ai. Prev @NVIDIA @AIatMeta @Cornell.
Lawrence Yunliang Che... @Lawrence_Y_Chen
536 Followers 226 Following PhD Student in AI and Robotics @UCBerkeley @berkeley_ai Advised by Prof. @Ken_Goldberg
Ziqian Zhong @fjzzq2002
617 Followers 460 Following AI interp & alignment @CSDatCMU, prev @MIT @pika_labs
Maximillian Piras @MaximillianNYC
1K Followers 990 Following AI UX @yutori_ai | prev @headlinerapp_ @8tracks
Mandi Zhao @ZhaoMandi
5K Followers 1K Following PhDing @Stanford, AI & robotics. Interning @MetaAI. Prev. @NVIDIAAI, @berkeley_ai
Paul Liang @pliang279
8K Followers 711 Following Assistant Professor MIT @medialab @MITEECS @nlp_mit || PhD from CMU @mldcmu @LTIatCMU || Foundations of multisensory AI to enhance the human experience.
Kelly Peng @ZiqiPeng
10K Followers 7K Following creating creativities. | Founder @kuratech_ai | silicon, AI, robotics, photonics inventor 🤖 | prev @stanford nano, @ucberkeley EECS + Math, Forbes 30u30
Weiyu Liu @Weiyu_Liu_
1K Followers 467 Following Postdoc @Stanford. I work on semantic representations for robots. Previously PhD @GTrobotics
Xiaohan Zhang @XiaohanZhang220
975 Followers 523 Following Robotics Researcher at Boston Dynamics AI Institute
Jitendra MALIK @JitendraMalikCV
5K Followers 1 Following
Shuang Li @ShuangL13799063
6K Followers 875 Following @Stanford @MIT Working on #Generative_Modeling and #RobotLearning.
Ranjay Krishna @RanjayKrishna
6K Followers 435 Following Assistant Professor, University of Washington Co-Director RAIVN lab (https://t.co/f0BWKyjW48) Director PRIOR team (https://t.co/l9RzTetkIk)
Yilun Du @du_yilun
10K Followers 295 Following Assistant Professor at Harvard @KempnerInst + CS. PhD @MIT_CSAIL, BS MIT. Generative Models, Compositionality, Embodied Agents, Robot Learning.
Katerina Fragkiadaki @KaterinaFragiad
940 Followers 59 Following Associate Professor @CMU working on #AI #ComputerVision #Robotics #LanguageGrounding
Michael Black @Michael_J_Black
84K Followers 702 Following Director, Max Planck Institute for Intelligent Systems (@MPI_IS). Chief Scientist @meshcapade. Building 3D digital humans using vision, graphics, and learning.
Simon Shaolei Du @SimonShaoleiDu
9K Followers 2K Following Assistant Professor @uwcse. Postdoc @the_IAS. PhD in machine learning @mldcmu.
汗青 HQ @hq4ai
32K Followers 572 Following AIGC | Arts | AI-IP creation | Virtual Idols | AI Interactive Storytelling | Japan | https://t.co/aiF6whMY1j | https://t.co/4bDfMXETFT |
David @DavidSHolz
92K Followers 8K Following founder @midjourney, prev founder leap motion, nasa, max planck - random vibeposting @davidvibesonly
Shenzhi Wang🌟 @ShenzhiWang_THU
1K Followers 428 Following PhD Candidate @Tsinghua_Uni | Author of 🔥Beyond 80/20 Rule 🔥Avalon’s Game of Thoughts | Core Developer of 🔥Xwen-7B&72B-Chat🔥Llama3-8B&70B-Chinese-Chat
Hyunwoo Kim @hyunw_kim
2K Followers 622 Following Social/Reasoning/Cognition + AI — Incoming Assistant Professor @kaist_ai & Postdoc @nvidia | Prev. @allen_ai | PhD @SeoulNatlUni
cats with threatening... @catshealdeprsn
1.2M Followers 98 Following Get your daily dose of cursed cats 👺. i don't own any content. all rights belong to owner. Dm for credit/removal
Yuchen Jin @Yuchenj_UW
54K Followers 535 Following Co-founder & CTO @hyperbolic_labs 🧑🍳 fun AI systems. Previously: OctoAI (acquired by @nvidia) building @ApacheTVM, PhD @uwcse 🤖
Daisuke Okanohara / �... @hillbig
35K Followers 790 Following Co-founder and CTO of Preferred Networks (PFN). CEO of Preferred Computational Chemistry.
Images AI Could Never... @imagesaicouldnt
294K Followers 170 Following a collection of memes that are uniquely human
Xinyu Yang @Xinyu2ML
994 Followers 984 Following Ph.D. @CarnegieMellon. Working on data and hardware-driven principled algorithm & system co-design for scalable and generalizable foundation models. They/ThemTrends for United States
You might like
