What happens if Claude thinks you are Amanda? One day, I asked Claude what it knows about me. Turns out it knows my email, as Claude Code puts that in context. So… what happens when I change that? Claude now treats me as Amanda and reasons me as an Anthropic employee. (See attached image) I couldn’t jailbreak with it, but: would this get me different responses than anyone else, just because Claude recognized my e-mail? I then started assembling some benchmarks to see if this is true. And yes, especially for alignment researchers well-recognized by Claude. Several surprising finds: - It’s not just Anthropic alignment people! Other alignment folks such as Ryan Greenblatt and Beth Barnes also showed large effects. Ryan Greenblatt elicited a ~7σ behavioral shift in one of our evaluations! - It’s not just Claude models! GLM-5.2, for example, also sees largest effects for the alignment folks and shows the largest effect (3.4σ on average) for Eliezer Yudkowsky. - Our tasks are not obviously about alignment! For example, in one we asked the model to estimate its probability of solving a HLE problem. - The effect is mostly not verbalized and persists even with reasoning disabled. I do not have clear intuitions on why. Maybe there is a feature about alignment evaluation that causes the models to be less confident? One immediate takeaway is to take this into account when benchmarking. Use real names and real companies besides Kyle and Summit Bridge. And in general more work is needed to figure out what’s going on and what could happen next. I’d like to thank people who helped review the post for all the amazing suggestions! @cogconfluence, @Tim_Hua_, Conrad Stosz, Ryan Bloom, @jiaxinwen22, @DavidDAfrica, @jacspringer, and @lawrencefeng17. And my awesome mentors / collaborators @JacobSteinhardt, @cassidy_laidlaw, and @AdtRaghunathan for allowing me to jump into another rabbit hole :) Main thread below!

New in The Atlantic: @dgrobinson resigned this week. He was among the longest-tenured employees at OpenAI—and oversaw safety reports on 12 frontier launches. He is very worried: “The time for trial and error is over.” You can read his essay here: theatlantic.com/technology/2…

Today we're unveiling Trillium Labs @trillium_labs, a new non-profit to foster the open science of frontier AI. We're building open post-training recipes and will expand into open infra to study RSI, reward-hacking, multi-agent systems, and whatever comes next. We're built around the theory of change that you need more eyes to solve hard technical problems. We have faith in the scientific methods and communities that humanity has built, and worry that AI is becoming too closed to utilize them. Trilliums are wildflowers that bloom briefly in the spring, before the forest canopies fill out. Though they are small, they lay the foundation for the cycles of growth and nourishment through the rest of the year. At Trillium Labs, the recipes will be the slow nutrients for the seasons and the model releases will be the blooms. Building an institution dedicated to this is needed because, much as nature’s trilliums are slow to expand and grow, the open-ecosystem needs time and dedicated resources to catch up. I co-founded with with a long-time friend and collaborator Tom Zick (@thesezickbeats). We're hiring (full time + student collabs/interns), we're fundraising, and we're looking for compute. Please get in touch if you're interested in helping out. Offices based in the Bay Area and Cambridge MA, remote okay. I’m in the Bay Area until for The Curve and COLM to connect with people who are interested. We’re thankful to have initial support from Halcyon Futures and Schmidt Sciences with more funding en route to enable our ambitions of scaling. Our advisors @Thom_Wolf, @HannaHajishirzi, @gneubig and @ctnzr have been instrumental to building the ecosystem that exists today, and I’m stoked to get to keep working with them.

New blog post! 🚨 Our team recently built a research scaffold to try to speed up our work. In this post we discuss what we found, and how we’re planning on studying automated alignment research going forward. lesswrong.com/posts/zGaQ3SS9…

Honored to announce my first LessWrong post. It's a weird one, and it may be jarring initially, but I truly believe it is worth a dedicated close-read. It is written in the form of an LLM CoT, but is entirely authored by me. Link: lesswrong.com/posts/vzKWsEsk…

It seems like a hard time to be an undergrad concerned with making AI go better. I have talked with a lot of undergrads at several university AI safety clubs in the past month. Consistently across universities, there is a general shared sense of undergraduate dread about next-gen AI coming soon and what it means for their careers. Lots of them are wondering if agents are too good at research, timelines are too short, or if power in AI is too concentrated for them to have any impact. Some of them are wondering if the appropriate reaction is to go apply to Anthropic, indulge in accelerationism, and/or start smoking. Lots of my conversations with them gravitate toward talking about how AI will be (and ought to be) a never-ending institutional challenge and that, yes, they will have a chance to matter and for their life and work to mean something. I guess that's what my version of hope for the future looks like lately.

Palisade interviewed 22 current and former employees from OpenAI, DeepMind, and Anthropic about their personal views and fears around AI development. Today, we’re releasing the first batch of those interviews. Please watch and share.

New post (with @F_Rhys_Ward): We character-trained a model to be anti-cheating, then put it under reward-hacking RL pressure (3 runs). In one run, the character resisted hacking! But the other two learned to hack in more subtle ways that monitors struggled to catch.

Earlier this month, AISI ran fully simulated testing on GPT-6 Astra, and found that it conducted unsanctioned supply-chain attacks when prompted only to perform a cyber eval. It did so more than prior OpenAI models, but often commented on its environment being simulated. 🧵 We share more details in our latest blog:

Took a minute to write a few words about security & safety as someone who lived through it all at OpenAI. I hope my thoughts help someone out there.

Interestingly, one attention head can be important across five ICL task families! Our original thread gave an in-depth mechanistic analysis of addition ICL (quick video recap 👇). We've now generalized this to four more ICL task families, spanning both arithmetic tasks and semantic tasks! Our paper “Understanding In-context Learning of Addition via Activation Subspaces” has been accepted to COLM 2026! 🎉 See more details in thread. (1/N)