Chi Jin
@chijinML
Followers
6K
Following
657
Media
24
Statuses
163
Associate Prof @Princeton. Previously: ML theory, RL & optimization. Now: AI for math, games & decision making.
Princeton, NJ
Joined November 2012
🚀 Huge milestone from our Goedel-Prover team: we’ve just released a new state-of-the-art model (8B & 32B) for automated theorem proving—surpassing the previous best 671B DeepSeek model by a wide margin, all with academic compute!
(1/4)🚨 Introducing Goedel-Prover V2 🚨 🔥🔥🔥 The strongest open-source theorem prover to date. 🥇 #1 on PutnamBench: Solves 64 problems—with far less compute. 🧠 New SOTA on MiniF2F: * 32B model hits 90.4% at Pass@32, beating DeepSeek-Prover-V2-671B’s 82.4%. * 8B > 671B: Our 8B
4
10
76
It was really nice talking to you all at such an interesting event. Special thanks to Seth and Jake for the tremendous effort in organizing!
The pokeagent workshop is tomorrow Featuring our speakers from Brown University and Google Deepmind, as well as announcing our competition winners Benchmarking in Pokemon is an exciting way to close-out NeurIPS!
0
2
16
Artificial intelligence demands massive new energy capacity. POW invests in the global companies building the backbone of this next-generation grid.
0
2
3
Join us at the 5th MATH-AI Workshop at @NeurIPSConf now, our biggest year yet with a record 249 submissions!! 🎉 ➡️ https://t.co/bLDwDZfWia We’re honored to host an incredible lineup of speakers: @swarat @WeizhuChen @j_dekoninck @Leonard41111588 @HannaHajishirzi @chijinML
Looking forward to seeing you at the MATH-AI workshop at #NeurIPS tomorrow! Location: Upper Level Ballroom 6A Schedule: https://t.co/OrkhDvseLY
4
12
54
At #NeurIPS2025? Join us for a Social on Wednesday at 7 PM, featuring a fireside chat with Jon Kleinberg and mentoring tables. Ft. mentors @canondetortugas @SurbhiGoel_ @HamedSHassani @tatsu_hashimoto @andrew_ilyas @chijinML @thegautamkamath @MountainOfMoon + more!
1
12
46
How do we close the gap between specialist RL and generalist LLM agents? We're benchmarking it in Pokémon. Join us at the PokeAgent Challenge competition workshop @ NeurIPS 2025. 📍 Dec 7, 8AM in San Diego 🎮 Track 1: Competitive Pokémon (game-theoretic reasoning) 🗺️ Track 2:
7
18
58
Super proud of my fantastic postdocs and graduate students taking their next steps at frontier labs 🎉 • Yong Lin (@Yong18850571) → Thinking Machine • Zihan Ding (@Hanry65960814) → Bytedance • Ahmed Khaled → Google It’s always bittersweet to say goodbye😢 but I couldn’t be
3
7
243
I am going to present two papers at #COLM2025 tomorrow from 4:30-6:30pm, as none of our leading authors can attend due to visa issues. Haven't done poster presentations for years 🤣🤣 .... so I will do my best! #76: LongProc #80: Goedel-Prover v1
Our Goedel-Prover V1 will be presented at COLM 2025 in Montreal this Wednesday afternoon! I won’t be there in person, but my amazing and renowned colleague @danqi_chen will be around to help with the poster — feel free to stop by!
4
26
347
Our Goedel-Prover V1 will be presented at COLM 2025 in Montreal this Wednesday afternoon! I won’t be there in person, but my amazing and renowned colleague @danqi_chen will be around to help with the poster — feel free to stop by!
2
8
73
Excited to share that I’ve been promoted to Associate Professor with tenure at Princeton!🎉 6 years may not be long, but AI research has evolved significantly during this period. Grateful to all my students, collaborators, colleagues for being with me on this remarkable journey!
150
64
3K
SCAP - a blended approach of capital appreciation and current income
1
2
36
🚀With early access to Tinker, we matched full-parameter SFT performance as in Goedel-Prover V2 (32B) (on the same 20% data) using LoRA + 20% of the data. 📊MiniF2F Pass@32 ≈ 81 (20% SFT). Next: full-scale training + RL. This is something that previously took a lot more effort
thinkingmachines.ai
How LoRA matches full training performance more broadly than expected.
2
18
186
Surprisingly, they achieve excellent results on Lean 4.15 by directly employing our prover, despite it being trained solely on Lean 4.9. With version alignment, the performance is expected to be even stronger.
0
0
12
I’m really amazed by this recent paper! It integrates Gemini 2.5 with our Goedel-Prover V2, redefining SOTA in formal mathematics and reaching new heights on PutnamBench, significantly beating even Seed-Prover! https://t.co/rE79C7MuNy
3
36
179
Welcome! 🎉 A special welcome to @chengshuai_shi , our new postdoc jointly hosted by my group and PLI, who will be leading exciting research on foundation models for games and decision making. Feel free to reach out if you’re interested!
This fall, we're welcoming 8 new postdocs! From reinforcement learning to human-AI collaboration, their work will power forward our initiatives. Meet them and learn more about their research: https://t.co/tfZwpgj72i
0
1
37
New paper up on arXiv! https://t.co/aiRJo6HVSp If you've used DiLoCo or other distributed optimization methods, you've seen this pattern: local optimizers on each node + an outer optimizer that combines updates. This always struck me as a bit strange, 1/7
arxiv.org
Modern machine learning often requires training with large batch size, distributed data, and massively parallel compute hardware (like mobile and other edge devices or distributed data centers)....
6
18
123
Don't miss out this wonderful event!
🔴 Final speaker lineup confirmed - PokéAgent Challenge Hackathon starts in 48 hours! NeurIPS 2025 competition featuring two tracks advancing AI decision-making through Pokémon: 🥊 Competitive battling 🏃 RPG speedrunning Research talks Saturday 12-1:30 PM EDT $2k in GCP
0
0
8
🔴 Final speaker lineup confirmed - PokéAgent Challenge Hackathon starts in 48 hours! NeurIPS 2025 competition featuring two tracks advancing AI decision-making through Pokémon: 🥊 Competitive battling 🏃 RPG speedrunning Research talks Saturday 12-1:30 PM EDT $2k in GCP
4
12
35
Our department (ECE) at Princeton is hiring in AI this year!📢 Please consider applying and joining us:
2
18
124
Thank you @KaiqingZhang for the kind invitation! Excited to speak at the Multiagent RL workshop at CDC in Brazil 🇧🇷---will be my first visit to both CDC and Brazil. Looking forward to it!
We have a fantastic lineup of speakers: Tamer Başar, @jababi, Rahul Jain, @chijinML, Cedric Langbort, Na Li, Aditya Mahajan, Prashant Mehta, @alexolshevsky1, Vijay Subramanian, and Serdar Yuksel. Please find more details at: https://t.co/QQ0lZTFjwY. Moreover, we also plan to
0
0
20
If you are going to @IEEECDC2025 this December in Rio 🇧🇷, please consider registering for the workshop I am helping co-organize with Aditya Mahajan and Vijay Subramanian, on Decentralized Information and Reinforcement Learning (RL), by Sept. 3rd at https://t.co/t3KxorRMsF! As
cdc2025.ieeecss.org
Registering for CDC 2025 secures your spot, grants access to sessions, networking, and resources, and ensures you're part of key industry insights.
1
2
17
In 2035, we’ll view unformalized 2025 math, where we don’t even name our hypotheses, in the same way that in 2025 we view pre 1600 algebra, before they realized they should name variables! Have you ever tried to read math pre 1600? They’re doing *serious* algebra, but it looks
33
68
457
📣 PostgreSQL developers: Meet Azure HorizonDB the new AI-ready database service with native AI capabilities built in. For enterprise scale, security, and performance without compromise.
1
5
41