berkeley_ai Profile Banner
Berkeley AI Research Profile
Berkeley AI Research

@berkeley_ai

Followers
234K
Following
383
Media
41
Statuses
1K

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

Berkeley, CA
Joined July 2017
Don't wanna be here? Send us removal request.
@ckwolfeofficial
C.K. Wolfe
3 days
New on @berkeley_ai blog by @seohong_park: divide-and-conquer value learning for off-policy RL. No TD bootstrapping. Scales to long horizons. https://t.co/VLspSfmRmW 🐻📄
Tweet card summary image
bair.berkeley.edu
The BAIR Blog
0
8
28
@2plus2make5
Emma Pierson
4 days
We don't fully understand the preferences human feedback encodes, so training on it can be risky. We propose a method to automatically discover these preferences! We identify unsafe, contradictory, and subjective preferences, and improve model safety, eval, and personalization.
@rajivmovva
Raj Movva
4 days
📣NEW PAPER! What's In My Human Feedback? (WIMHF) 🔦 Human feedback can induce unexpected/harmful changes to LLMs, like overconfidence or sycophancy. How can we forecast these behaviors ahead of time? Using SAEs, WIMHF automatically extracts these signals from preference data.
2
10
86
@truthlibertyco
Truth & Liberty
2 months
Does faith belong in the public square? At Truth & Liberty, we say yes. Follow us for bold conversations about faith, freedom, and family values.
2
1
19
@rajivmovva
Raj Movva
4 days
📣NEW PAPER! What's In My Human Feedback? (WIMHF) 🔦 Human feedback can induce unexpected/harmful changes to LLMs, like overconfidence or sycophancy. How can we forecast these behaviors ahead of time? Using SAEs, WIMHF automatically extracts these signals from preference data.
8
35
209
@dawnsongtweets
Dawn Song
4 days
Honored to be selected as an #AI2050 Senior Fellow by @SchmidtSciences, alongside inspiring researchers shaping AI for the benefit of humanity. Excited to continue advancing the work on AI, AI Safety &  Security — building AI-powered systems that are safer, more trustworthy,
@schmidtsciences
Schmidt Sciences
5 days
We're excited to welcome 28 new AI2050 Fellows! This 4th cohort of researchers are pursuing projects that include building AI scientists, designing trustworthy models, and improving biological and medical research, among other areas. https://t.co/8oY7xdhxvF
5
13
101
@JosephJSSuh
Joseph Jeesung Suh
5 days
LLMs have dominated recent work on simulating human behaviors. But do you really need them? In discrete‑choice settings, our answer is: not necessarily. A lightweight graph neural network (GNN) can match or beat strong LLM-based methods. Paper: https://t.co/WvMRy4DdjR 🧵👇
3
14
53
@kvfrans
Kevin Frans
10 days
What really matters in matrix-whitening optimizers (Shampoo/SOAP/PSGD/Muon)? We ran a careful comparison, dissecting each algorithm. Interestingly, we find that proper matrix-whitening can be seen as *two* transformations, and not all optimizers implement both. Blog:
5
48
326
@kumarkagrawal
kk
11 days
Autoregressive language models learn to compress data by mapping sequences to high-dimensional representations and decoding one token at a time. The quality of compression, as defined by the ability to predict the next token given a prompt, progressively improves (as measured by
4
22
79
@ameeshsh
Ameesh Shah
12 days
LLMs have shown a remarkable ability to “self-refine” and learn from their mistakes via in-context learning. But in robotics, most methods are single-shot. How can we bring inference-time adaptation to robot learning? A 🧵:
10
18
129
@svlevine
Sergey Levine
12 days
New work from @aditya_oberai & @seohong_park: instead of 1-step TD backups or n-step, can we "divide and conquer" over the trajectory, backing up finer and finer increments? Improves over bias of TD-0 and variance of MC. Principle is old, but getting it to work takes some care!
@aditya_oberai
Aditya Oberai
12 days
TD Learning can suffer on long tasks: ↑ deep bellman recursions ��� ↓ poor scalability (despite big data) We introduce a new method (TRL) with a "divide-and-conquer" value update, which scales well with long horizons!
6
27
244
@aditya_oberai
Aditya Oberai
12 days
TD Learning can suffer on long tasks: ↑ deep bellman recursions → ↓ poor scalability (despite big data) We introduce a new method (TRL) with a "divide-and-conquer" value update, which scales well with long horizons!
2
29
231
@BaruaJosh
Josh Barua
13 days
🌍 LLMs can use long chain-of-thought (CoT) to reason in English, but what about other languages? New paper w/ @BerkeleyNLP: We study how scaling, pretraining, post-training & inference affect long CoT across 9 languages. Spoiler: English long CoT ≠ multilingual long CoT 🧵
3
8
22
@_ahmedmalaa
Ahmed Alaa
14 days
Our new paper with Sonali Sharma and @RoxanaDaneshjou is out in @npjDigitalMed! We examine how medical safety and disclaimer messages in public LLMs have changed over time when answering patient questions.
@npjDigitalMed
npj Digital Medicine
20 days
Generative AI models are giving fewer medical disclaimers over time. 📉 In 2022, ~26% of AI health answers had a disclaimer. By 2025? <1%. As models get smarter, they’re getting less safe. Patients may take outputs as medical advice. https://t.co/2OYQvKdezT
3
8
17
@henseoba
Wen-Han Hsieh
16 days
AI can now see, reason, and segment the Earth. 🌍 Meet LISAt, our #NeurIPS2025 Datasets & Benchmarks paper - the first foundation model that turns language queries into pixel-level satellite segmentations. 🛰️ (1/n) 🔗 https://t.co/ApVZgGF0cU @NeurIPSConf @berkeley_ai
4
3
29
@ZehanMa123
Zehan Ma
18 days
Can a robot inspect all views of an object? Today @IROS, we present Omni-Scan from @berkeley_ai, a novel method for bimanual robo 360° object scanning & reconstruction using 3D Gaussian Splats. (1/8) 🔗 https://t.co/8emyJfUNk4
5
12
122
@akshatgupta57
Akshat Gupta
19 days
🧠 New preprint: How Do LLMs Use Their Depth? We uncover a “Guess-then-Refine” mechanism across layers - early layers predict high-frequency tokens as guesses; later layers refine them as context builds Paper - https://t.co/5PitHjmJJZ @neuranna @GopalaSpeech @berkeley_ai
15
73
521
@sfiscience
Santa Fe Institute
19 days
Catch up on our most recent Community Lecture: “Transmission Versus Truth: What Will It Take to Make an AI as Smart as a 4-Year-Old?” with Alison Gopnik.  This was the last of six community lectures for 2025, and all are available to watch on SFI’s YouTube channel. Watch here:
1
12
27
@dawnsongtweets
Dawn Song
20 days
New evaluation results from @AnthropicAI's Claude Sonnet 4.5’s system card on our CyberGym benchmark reveals a striking trend: AI cybersecurity capabilities are advancing at unprecedented speed—from ~10% (Claude-Sonnet -3.7) to ~30% success rates (Claude-Sonnet-4.5) (with single
@dawnsongtweets
Dawn Song
5 months
1/ 🔥 AI agents are reaching a breakthrough moment in cybersecurity. In our latest work: 🔓 CyberGym: AI agents discovered 15 zero-days in major open-source projects 💰 BountyBench: AI agents solved real-world bug bounty tasks worth tens of thousands of dollars 🤖
7
15
56
@Berkeley_EECS
UC Berkeley EECS
20 days
Amazing! 10 @BerkeleyEECS @SkyCompLab grad students are Amazon AI PhD Fellows! Congrats! Learn more about our fellows here: https://t.co/zuCGKlmSNe #AmazonAIFellowship @BerkeleySky
Tweet card summary image
eecs.berkeley.edu
Today, Amazon announced its new AI PhD Fellowship program, offering two years of funding to over 100 PhD students across nine universities. Ten of these inaugural fellowships have been awarded to...
@AmazonScience
Amazon Science
20 days
🎓 Amazon launches AI PhD Fellowship program, providing $68 million over two years to fund PhD students at 9 universities pursuing research in machine learning, computer vision, and natural-language processing. #AmazonAIFellowship
0
14
59
@tsunghan_wu
Tsung-Han (Patrick) Wu
20 days
Humans handle dynamic situations easily, what about models? Turns out, they break in three distinct ways: ⛔ Force Stop → Reasoning leakage (won’t stop) ⚡️ Speedup → Panic (rushed answers) ❓ Info Updates → Self-doubt (reject updates) 👉Check out https://t.co/wKrnsMkiFY
5
20
67