• Home
• Projects and Writing
• About

Projects and Writing

A selection of works I am proud of. These were done with many wonderful collaborators.

• 2024-09-01 — 2025-10-24   Learning to interpret weight differences in language models
• 2023-10-12 — 2024-07-01   Covert malicious finetuning
• 2023-07-20 — 2024-06-20   Can Go AIs be adversarially robust?
• 2023-06-08 — 2023-12-15   Takeaways from a mechinterp project
• 2021-09-15 — 2023-07-20   Adversarial policies beat superhuman Go AIs
• 2021-01-10 — 2021-07-02   Taylor expansions: An easy derivation
• 2015-06-01 — 2015-12-23   Codeforces round #336