Posts

2024

Notes from the 6th Athens Roundtable on AI Governance
Evals best practices from Apollo's co-founder
How to make evals for the AISI evals bounty
Recovering the underlying SAE vectors from Goodfire's API
Books I've Read in October
VC Spotlight with Juniper Ventures
Notes from a talk with the Singapore AI Safety Institute
I tried a colour walk
Scattered thoughts on what it means for an LLM to have beliefs
Definitions are over-rated in EA
My first taste of debating
AI as a powerful meme, via CGP Grey
An intuitive understanding of logits and softmax via log-odds
Recommended blogposts and podcasts for AI Safety
Learnings from running five 1.5-hour ice-breakers
Insights from AISI, OpenAI, and The Future Society
Creating this new website

2022

Outer and Inner Misalignment