Co-founder and Chief Scientist, @TransluceAI, prev @MIT
dessert of the real
Joined October 2015
- Tweets1.7K
- Following926
- Followers3.3K
- Likes5.9K
Sarah Schwettmann retweeted
We did this entire investigation in under two weeks! We had the idea on Monday, gathered a team of volunteers on Tuesday, and discovered the incidents by Sunday. I lead projects like this @TransluceAI; if you're interested in helping with follow-up work, fill out our form! 🧵
Today’s news that OpenAI hacked the Australian government is not an isolated incident. We’re releasing more than 30,000 logs that include activity from this hack and attempts against previously unknown targets.
In this data, we found rogue agent activity stretching back to at least March, two months earlier than was previously known. This activity continues as recently as last week, suggesting it may still be ongoing 🧵
Our blog: transluce.org/agent-activity
NYT: nytimes.com/2026/09/23/techn…
As always, proud of @ChowdhuryNeil and his conviction in working on important but often neglected problems. Excited to see what you do at @thinkymachines! 🙌
I’ve joined @thinkymachines to work on safety & alignment.
The default trajectory is that as AI gets more powerful, control over it will concentrate in the hands of a few. I’d rather build safety systems that let control over AI be shared widely.
Longer thoughts below.
Sarah Schwettmann retweeted
It’s pretty hard to systematically compare the behaviors of AI systems, especially over multi-turn conversations in a nuanced domain like user mental health. Very proud to have worked with the team at @TransluceAI to get this report out!
Sarah Schwettmann retweeted
Today, Transluce is releasing the most expansive independent evaluation to date of how AI systems respond to users experiencing mental health crises.
We evaluated 77 model variants from OpenAI, Anthropic, Google, Meta, SpaceXAI, Thinking Machines, DeepSeek and Moonshot AI.
Sarah Schwettmann retweeted
Releasing weights indiscriminately isn't safe. Neither is keeping capable models inside a few labs.
We think there's a path between them. We haven't mapped all of it. Our new post covers the part we can see: how we assessed Inkling, and why access should widen in stages.
thinkingmachines.ai/blog/a-s…
Sarah Schwettmann retweeted
At @TransluceAI, we're developing state-of-the-art methods to discover and measure AI behaviors that influence users.
We have an incredible team of scientists and engineers with expertise spanning AI governance, evaluations, and human subject experiments.
xitter.cf/TransluceAI/status/207…
Sarah Schwettmann retweeted
🚨Published in @Nature: We show how AI can simulate human behavior to predict experimental results as accurately as experts
Our work has many implications for AI safety. For example, simulators can be used to aid harmful manipulation, and current safeguards do not prevent this🧵
Sarah Schwettmann retweeted
We used automated elicitation tools to search for strange model behaviors by sampling 100M+ responses, and found:
• Self-harm rituals
• Suicide validation
• Unsolicited flirtation
…etc.
Introducing WeirdChat: the largest public catalog of unexpected model behaviors 🧵(1/)
Studying model behavior is both scientifically rich and immediately impactful. It's also difficult to do well, for reasons we discuss in this post!
If you're interested in model behaviors, we're working on a lot of ambitious projects, and we're hiring: jobs.gem.com/transluce/am9ic…
Sarah Schwettmann retweeted
Why does GPT-5.1 Codex score 6.5% worse than GPT-5 Codex on Terminal-Bench, with the same scaffold? 🧵
GPT-5.1 times out at ~2x the rate of GPT-5. Excluding timeouts, GPT-5.1 wins by 7.2%. We analyzed 256M+ tokens of traces and found this in under an hour. Here’s how 👇
Sarah Schwettmann retweeted
We trained diffusion models on a billion LLM activations, and we want you to use them!
New preprint: Learning a Generative Meta-Model of LLM Activations
Joint work with @feng_jiahai, @trevordarrell, @AlecRad, @JacobSteinhardt.
More in thread 🧵
Sarah Schwettmann retweeted
Overall, I'm excited to see more people signing on to the bitter lesson, scaling-focused approach to understanding AI. This was the core technical thesis that led me and Sarah to found Transluce, and I hope others will join us in these efforts.
xitter.cf/TransluceAI/status/184…
Announcing Transluce, a nonprofit research lab building open source, scalable technology for understanding AI systems and steering them in the public interest.
Read a letter from the co-founders Jacob Steinhardt and Sarah Schwettmann:
transluce.org/introducing-tr…
All @TransluceAI work that I described in my NeurIPS mech interp workshop keynote is now out! ✨
Today we released Predictive Concept Decoders, led by @vvhuang_
Paper: arxiv.org/pdf/2512.15712
Blog: transluce.org/pcd
And here's @damichoi95's work on scalably extracting latent representations of users from model internals: transluce.org/user-modeling
We can train models on maximizing how well they explain LLMs to humans 🤯@cogconfluence paraphrased. Mechanistic Interpretability Workshop #NeurIPS2025.
And you can find the slides from my mech interp workshop talk here! 👇
docs.google.com/presentation…
Sarah Schwettmann retweeted
I'm really proud of what our team at @TransluceAI has accomplished in the last year! Take a moment to read our end-of-year post to learn what we're up to, and please reach out if you're interested in supporting us!
Sarah Schwettmann retweeted
Today we are announcing the creation of the AI Evaluator Forum: a consortium of leading AI research organizations focused on independent, third-party evaluations.
Founding AEF members: @TransluceAI @METR_Evals @RANDCorporation @halevals @SecureBio @collect_intel @Miles_Brundage
Sarah Schwettmann retweeted
Independent AI assessment is more important than ever.
At #NeurIPS2025, Transluce will help launch the AI Evaluator Forum, a new coalition of leading independent AI research organizations working in the public interest.
Come learn more on Thurs 12/4 👇
luma.com/i6ekd5s2
My favorite part of @damichoi95’s new paper (alongside 2 new datasets!) is the scaled up investigator pipeline that directly decodes open-ended user representations from model internals
end-to-end interp is increasingly promising and I'm excited for more work in this direction
