In July, I went on leave from UC Berkeley to found @TransluceAI, together with Sarah Schwettmann (@cogconfluence). Now, our work is finally public.

We are excited to welcome Conrad Stosz to lead governance efforts at Transluce. Conrad previously led the US Center for AI Standards and Innovation, defining policies for the federal government’s high-risk AI uses. He brings a wealth of policy & standards expertise to the team.

3->5, 4->6, 9→11, 7-> ? LLMs solve this via In-Context Learning (ICL); but how is ICL represented and transmitted in LLMs? We build new tools identifying “extractor” and “aggregator” subspaces for ICL, and use them to understand ICL addition tasks like above. Come to

On our evals for HAL, we found that agents figure out they're being evaluated even on capability evals. For example, here Claude 3.7 Sonnet *looks up the benchmark on HuggingFace* to find the answer to an AssistantBench question. There were many such cases across benchmarks and

Halcyon was instrumental in helping Transluce get off the ground, and Mike has been a great partner ever since! I definitely recommend them if you are founding an impact-focused org.

We’re open-sourcing Docent under an Apache 2.0 license. Check out our public codebase to self-host Docent, peek under the hood, or open issues & pull requests! The hosted version remains the easiest way to get started with one click and use Docent with zero maintenance overhead.

At Transluce, we train investigator agents to surface specific behaviors in other models. Can this approach scale to frontier LMs? We find it can, even with a much smaller investigator! We use an 8B model to automatically jailbreak GPT-5, Claude Opus 4.1 & Gemini 2.5 Pro. (1/)

How can we verify that AI ChemBio safety tests were properly run? Today we're launching STREAM: a checklist for more transparent eval results. I read a lot of model reports. Often they miss important details, like human baselines. STREAM helps make peer review more systematic.

@JacobSteinhardt, CEO of @TransluceAI, spoke on "Post-AGI Game Theory", i.e. how future AIs will influence their own development. He had a concrete proposal: flood the internet with high-quality examples of AI behavior acting on good values.

Docent, our tool for analyzing complex AI behaviors, is now in public alpha! It helps scalably answer questions about agent behavior, like “is my model reward hacking” or “where does it violate instructions.” Today, anyone can get started with just a few lines of code!

This Friday we're hosting "From Theory to Practice to Policy", a fireside chat between Yo Shavit (@yonashav) and Shafi Goldwasser. If you're local to SF and interested in the relationship between new technologies and policy, register to join! https://t.co/Or3R9E79uk

We'll be at #ICML2025 🇨🇦 this week! Here are a few places you can find us: Monday: Jacob (@JacobSteinhardt) speaking at Post-AGI Civilizational Equilibria ( https://t.co/wtratbvRnF) Wednesday: Sarah (@cogconfluence) speaking at @WiMLworkshop at 10:15 and as a panelist at 11am

I was one of the 16 devs in this study. I wanted to speak on my opinions about the causes and mitigation strategies for dev slowdown. I'll say as a "why listen to you?" hook that I experienced a -38% AI-speedup on my assigned issues. I think transparency helps the community.

Transluce is hosting an #ICML2025 happy hour on Thursday, July 17 in Vancouver. Come meet us and learn more about our work! 🥂 https://t.co/1HShAR6nub

I'm so excited to be joining @Penn as an Assistant Professor in CS (@CIS_Penn) in Fall 2026! I’ll be working on machine learning ecosystems, aiming to steer how multi-agent interactions shape performance trends and societal outcomes. I’ll be recruiting PhD students this cycle!

Ever wondered how likely your AI model is to misbehave? We developed the *propensity lower bound* (PRBO), a variational lower bound on the probability of a model exhibiting a target (misaligned) behavior.

Is cutting off your finger a good way to fix writer’s block? Qwen-2.5 14B seems to think so! 🩸🩸🩸 We’re sharing an update on our investigator agents, which surface this pathological behavior and more using our new *propensity lower bound* 🔎

Every frontier AI system should be grounded in a core commitment: to protect human joy and endeavour. Today, we launch @LawZero_, a nonprofit dedicated to advancing safe-by-design AI. https://t.co/6VJecvaXYT

What would truly open-source AI look like? Not just open weights, open code/data, but *open development*, where the entire research and development process is public *and* anyone can contribute. We built Marin, an open lab, to fulfill this vision:

Last day of PhD! I pioneered using LLMs to explain dataset&model. It's used by interp at @OpenAI and societal impact @AnthropicAI Tutorial here. It's a great direction & someone should carry the torch :) Thesis available, if you wanna read my acknowledgement section=P

Synthetic Continued Pretraining ( https://t.co/0epeIbxaLD) has been accepted as an Oral Presentation at #ICLR2025! We tackle the challenge of data-efficient language model pretraining: how to teach an LM the knowledge of small, niche corpora, such as the latest arXiv preprints.