October AI Safety Paper Highlights: Clean Neurons, OpenAI Monitoring, Natural Schemers
gasteigerjo · x · 2026-11-04
Johannes Gasteiger's monthly roundup of AI safety research highlights covers: clean neurons, OpenAI's monitoring and incident processes, self-deployment on IoT devices, simulated control evals, routing reward hacks, natural schemers, and inheriting misalignment. Full list on his AI Safety at the Frontier Substack.
More from Safety
- Margaret Mitchell: Weight Communication Without Human Decoding Is the Wrong Path — mmitchell_ai · 2026-11-04
- Mitchell: Fixing Weight-Communication Risks Is Counted as "Slowing Down" — mmitchell_ai · 2026-11-04
- Postman desktop auto-installs Codex/Claude plugins, tries to trust its own startup hook — jedisct1 · 2026-11-04
- Google DeepMind researchers say contracts bar them from public AI safety debate — mmitchell_ai · 2026-11-04
- Put static secrets in your HTTP proxy: exe.dev on agent-era key management — davidcrawshaw · 2026-11-04
- Frontier AI Safety and Governance 2026 report released at Wuzhen Summit, mapping AI risks — pstAsiatech · 2026-11-04