Member of Technical Staff @AnthropicAI CodeRL/Alignment | PhD @ColumbiaCompSci | NLP & ML | Prev Intern @MSFTResearch, @AmazonScience
San Francisco, CA
Joined January 2019
- Tweets107
- Following609
- Followers2.6K
- Likes653
Yanda Chen retweeted
From OSWorld 1.0 to 2.0, we went from minutes (~30 steps) to hours (~318), from single apps to real workflows, from high scores (83%) to hard problems (21%).
1+ year, 20+ people, every task rigorously verified. This is what real cua evaluation takes.🙏
👉osworld-v2.xlang.ai
Two years ago, we built OSWorld 1.0 — the benchmark that became the standard for computer-use agents. Agents now score 83.5% on it. Problem solved?
Not even close.
🚀Today we introduce OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks.
What's new:
🎯 108 real-world workflows, each ~1.6 hours ⏱️ for a skilled human
⚙️ ~318 tool calls/task vs. ~30 in OSWorld 1.0
🌍 Grounded in authentic artifacts & stateful user profiles
⚡ Captures real phenomena: dynamic environments, streaming interaction, cross-source reasoning, implicit-state inference & more
📊 Best results: Claude Opus 4.8 reaches the highest accuracy at 20.6%, while GPT-5.5 is far more token-efficient but plateaus near 13%. No one is close to solving real computer use.
🏠 Homepage: osworld-v2.xlang.ai
📄 Paper: github.com/xlang-ai/OSWorld-…
💻 Code: github.com/xlang-ai/OSWorld-…
🤗 Dataset: huggingface.co/datasets/xlan…
🧵 [1/8]
Yanda Chen retweeted
Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software.
It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans.
anthropic.com/glasswing
Yanda Chen retweeted
Could an LLM have emotions? It’s hard to say. But when you’re talking to Claude, ChatGPT, or Gemini, you’re not talking to an LLM. You’re talking to a *character* being authored by an LLM. And these characters can, functionally, be driven by internal representations of desperation, or fear, or empathy (with sometimes alarming consequences).
Excited that the last work back from my PhD is out! We proposed a method to train language models to control the factuality–informativeness trade-off in their responses based on user preferences.
@SaraZiweiGong has been working on a lot of interesting stuff at the intersection of AI and psychology — worth following if you’re interested in those areas!
(1/4) Excited to share our new preprint: "Factuality on Demand"with @yanda_chen_ !
LLMs face an inherent trade-off: be cautious but uninformative, or detailed but prone to hallucinations. What if you could tune this with a simple knob? Read: arxiv.org/abs/2602.00848
#NLP #LLMs
Yanda Chen retweeted
(1/6) Why does next-token prediction work so well, even for long text? 🤔
Check out “Provable Long-Range Benefits of Next-Token Prediction”.
A rigorous explanation for LLM’s long-range coherence/reasoning.
Joint work with Santosh Vempala📄 arXiv: arxiv.org/abs/2512.07818
Yanda Chen retweeted
Reward hacking means the model is making less effort than expected: it finds the answer long before its fake CoT is finished. TRACE uses this idea to detect hacking when CoT monitoring fails. Work led by @XinpengWang_ @nitishjoshi23 and @rico_angell👇
This Post is from an account that no longer exists. Learn more
Yanda Chen retweeted
We’re hiring someone to run the Anthropic Fellows Program!
Our research collaborations have led to some of our best safety research and hires. We’re looking for an exceptional ops generalist, TPM, or research/eng manager to help us significantly scale and improve our collabs 🧵
Yanda Chen retweeted
We recently ran to have OpenAI and Anthropic each evaluate each others’ models for safety issues. Excited for us to find more ways to help support safety practices across the whole field!
Our results overall suggest that we can effectively separate harmful from harmless data and use pretraining data filtering to improve model safety without compromising usefulness.
Big thanks to the team! 🙏
@MycalTucker, @NinaPanickssery, @marsupialtail_2, @framosconis, @anjali_gopal, Carson Denison, @petrini_linda
@janleike, @EthanJPerez, @MrinankSharma
New Anthropic research: filtering out dangerous information at pretraining.
We’re experimenting with ways to remove information about chemical, biological, radiological and nuclear (CBRN) weapons from our models’ training data without affecting performance on harmless tasks.
ALT Pretraining data filtering pipeline. We automatically scored the harmfulness of each document in a model’s pretraining dataset with a classifier and removed those above a certain threshold. We then pretrained the model from scratch on the filtered dataset containing only harmless content (as determined by the classifier).
Yanda Chen retweeted
If you want to get into alignment research, imo this is one of the best ways to do it.
Some previous fellows did some of the most interesting research I've seen this year and >20% ended up joining Anthropic full-time.
Application deadline is this Sunday!
We’re running another round of the Anthropic Fellows program.
If you're an engineer or researcher with a strong coding or technical background, you can apply to receive funding, compute, and mentorship from Anthropic, beginning this October. There'll be around 32 places.
ALT A drawing of two hands manipulating abstract shapes
Yanda Chen retweeted
After we published our circuit tracing work, researchers from several different interpretability groups came together for this collaboration. Check it out for replications, extensions, and perspectives on the field, including suggestions for future work!
Today, we're releasing The Circuit Analysis Research Landscape: an interpretability post extending & open sourcing Anthropic's circuit tracing work, co-authored by @Anthropic, @GoogleDeepMind, @GoodfireAI @AiEleuther, and @decode_research.
Here's a quick demo, details follow: ⤵️
Yanda Chen retweeted
We're doubling the size of Anthropic's Fellows Program and launching a new round of applications.
The first round of collaborations led to a number of recent/upcoming safety results that are comparable in impact to work our internal safety teams have done (IMO)
We’re running another round of the Anthropic Fellows program.
If you're an engineer or researcher with a strong coding or technical background, you can apply to receive funding, compute, and mentorship from Anthropic, beginning this October. There'll be around 32 places.
ALT A drawing of two hands manipulating abstract shapes
So excited to hear that Kathy won the ACL Lifetime Achievement Award! I feel incredibly fortunate and honored to have had her as one of my PhD advisors, and I’ve learned so much from her over the years. Big congrats!
🕊️ Lifetime Achievement Award at #ACL2025NLP
A standing ovation for Prof. Kathy McKeown, recipient of the ACL 2025 Lifetime Achievement Award! 🌟
Yanda Chen retweeted
New Anthropic Research: “Inverse Scaling in Test-Time Compute”
We found cases where longer reasoning leads to lower accuracy.
Our findings suggest that naïve scaling of test-time compute may inadvertently reinforce problematic reasoning patterns.
🧵
Yanda Chen retweeted
📰We've just released SHADE-Arena, a new set of sabotage evaluations. It's also one of the most complex, agentic (and imo highest quality) settings for control research to date!
If you're interested in doing AI control or sabotage research, I highly recommend you check it out.
My first paper @AnthropicAI is out!
We show that Chains-of-Thought often don’t reflect models’ true reasoning—posing challenges for safety monitoring.
It’s been an incredible 6 months pushing the frontier toward safe AGI with brilliant colleagues. Huge thanks to the team! 🙏
@JoeJBenton, @anshrad, @JonathanUesato, Carson Denison, @johnschulman2, Arushi Somani,
@peterbhase, @MishaWagne29322, @FabienDRoger, Vlad Mikulik, @sleepinyourhat, @janleike, Jared Kaplan, @EthanJPerez
🔗 assets.anthropic.com/m/71876…
New Anthropic research: Do reasoning models accurately verbalize their reasoning?
Our new paper shows they don't.
This casts doubt on whether monitoring chains-of-thought (CoT) will be enough to reliably catch safety issues.
ALT Title card for the paper "Reasoning Models Don't Always Say What They Think", by Chen et al.
Yanda Chen retweeted
My favorite figure from our new Circuits papers -- "How does Claude do math?"
Claude simultaneously does:
1. a back of the envelope calculation of the tens digits -- "the answer should be somewhere around 90".
2. an exact calculation of 6+9=15 using these super cool look up table features.
Yanda Chen retweeted
Human thought is built out of billions of cellular computations each second.
Language models also perform billions of computations for each word they write. But do these form a coherent “thought process?”
We’re starting to build tools to find out! Some reflections in thread.
Yanda Chen retweeted
New Anthropic blog post: Subtle sabotage in automated researchers.
As AI systems increasingly assist with AI research, how do we ensure they're not subtly sabotaging that research? We show that malicious models can undermine ML research tasks in ways that are hard to detect.
Yanda Chen retweeted
Could we spot a misaligned model in the wild?
To find out, we trained a model with hidden misalignments and asked other researchers to uncover them in a blind experiment.
3/4 teams succeeded, 1 of them after only 90 min
