Writing
2026
- GPT-6 Astra can do a lot of multi-hop reasoning without chain of thought
- Perspectives on Continual Learning: Survey Results and Forecasts
- Angles of attack for continual learning safety
- How might continual learning affect safety and alignment?
- What's Continual Learning, and Why Might We Expect To See It In Advanced LLM Agents?
- Implications of Continual Learning for LLM Agents: Introduction
- Should We Train Against (CoT) Monitors?
- [Paper] How does information access affect LLM monitors' ability to detect sabotage?
- A quick, elegant derivation of Bayes' Theorem
- Exploring Reinforcement Learning Effects on Chain-of-Thought Legibility
- Aether is hiring technical AI safety researchers
2025
2024