Berkeley AI Research

This account is verified. Learn more

@berkeley_ai

We're graduate students, postdocs, faculty and scientists at the cutting edge of artificial intelligence research.

Berkeley, CA bair.berkeley.edu Joined July 2017

491 Following 301K Followers 1.7K Posts

Latest posts 5

Berkeley AI Research @berkeley_ai Oct 6

Congratulations to @berkeley_ai faculty @irenetrampoline and @berkeley_ai alumni @BachFrancis @andrea_bajcsy @dan_fried @natolambert Juan Carlos Perdomo who have been named @schmidtsciences 2026 AI2050 Early Career Fellows.

eecs.berkeley.edu/news/irene-che…?

eecs.berkeley.edu Irene Chen and 6 EECS alumni named Schmidt Sciences AI2050 Fellows - EECS at Berkeley UC Berkeley EECS Prof. Irene Chen and 6 alumni were named 2026 Schmidt Sciences AI2050 Fellows to tackle critical long-term AI challenges. Read more.

Berkeley AI Research @berkeley_ai Oct 3

Congratulations to the @berkeley_ai Center for Humanoid Intelligence team members who placed 1st at the two tracks in the Robotic Collaborative Assembly Challenge at #IROS2026!

Wei ZHAN @Wei_ZHAN_ Oct 2

Proud of my team members who won the 1st place of two tracks in the RoCo Challenge at #IROS2026 in Pittsburgh!

Robotic Collaborative Assembly Challenge (rocochallenge.github.io/RoCo-IROS2026/) includes challenging tasks like industrial board and brick assembly on bimanual and tactile-aware setups with 55 registered teams from 8+ countries.

In addition to days of hard work via onsite hackathon, the winning point also resides in our recent work at BAIR Center for Humanoid Intelligence (HIC) leveraging simulation, which will be released in the future.

Kudos to the team Jinzhong Gu, Max Wang, @FanXu22, Xingjian Wang, @MutianShen, @jiakai_chen3, Xinyi Cao, Jingdong Zhou, Chenrui Tie, Haoran Lu, Shuo Yin, from @berkeley_ai and MSC Lab at UC Berkeley, with nice guidance from @warshallrho and Masayoshi Tomizuka!

Kudos to the team Jinzhong Gu, Max Wang, @, Xingjian Wang, e organizers @RuixuanLiu_, Haichao Liu, @ZiweiWangNTU and @ChangliuL!@

View quoted post on X

Berkeley AI Research @berkeley_ai Sep 29

Presenting research from @berkeley_ai Humanoid Intelligence Center.

DexTacWAM: A Visuo-Tactile World-Action Model for Dexterous Manipulation.🖐️🤖

Dexterous manipulation requires touch, yet multi-finger tactile data remain scarce and expensive to collect at scale.

Through continual vision-to-touch learning, DexTacWAM adapts a pretrained video world model into a visuo-tactile world model for predictive multi-finger contact modeling and action generation.

Across six contact-rich dexterous manipulation tasks, DexTacWAM achieves the highest score on every task, averaging 70.6 vs. 38.0 for the strongest baseline.

🔬 Why not simply inject tactile features into the action policy?

With the same multi-finger tactile encoder, policy architecture, and training procedure, replacing the tactile world-model latent with direct tactile features drops the 4-task mean from 74.7 to 26.6.

Takeaway: the benefit does not come simply from providing tactile observations to the policy, but from making multi-finger contact evolution part of the predicted world state.

All code, model weights, and datasets are now open-sourced!

🌐 Project: dextacwam.github.io
📄 Paper: arxiv.org/abs/2609.24976
💻 Code: github.com/dextacwam/DexT…
🤗 Weights&Data: huggingface.co/collections/Je…

Author List: @Jensen_Yuan, @zekaiw04, @sbn_epiphany Haoran Lu, @trevordarrell, @Ismini_L, @Wei_ZHAN_

2:25

Berkeley AI Research @berkeley_ai Jul 24

@berkeley_ai faculty Peter Bartlett is giving a plenary lecture at this year's International Congress of Mathematicians #ICM2026 on July 28, 2026
11:30 AM-12:30 PM ET.

His lecture "Modern Machine Learning Methods: Implicit Bias, Benign Overfitting, and Unstable Optimization" is online here: epubs.siam.org/doi/10.1137/25…

epubs.siam.org Proceedings of the International Congress of Mathematicians 2026 - Volume 2: Plenary Lectures |... Abstract. The impressive performance of modern machine learning methods seems to arise through different mechanisms from those of classical statistical learning theory, mathematical statistics, and...

Berkeley AI Research @berkeley_ai Jul 1

We could not be prouder of this years BAIR PhD graduates! Follow their journey to see where they go next @baifeng_shi @sea_snell @databoydg @enfleisig @graceluo_ @zhuhl98 @HaozhiQ @jdzamfi @LianJiachen @joshminwookang @junhaobearxiong @kaylolittlejohn @KentKChang @kvablack @KunheYang @lisabdunlap @LongTonyLian @maulik__bhatt @michaelpsenka @nathanlichtle @neerjathakkar @nikita_mehandru @NiklasLauffer @qiyang_li @vbenara05 @vonekels @xiuyu_l @Yiheng_Li_Cal @ZheFu1 Sampada Deglurkar Wei-Jer Chang Yichen Xie Yigit Erginbas

bair.berkeley.edu/blog/2026/07/0…