Redwood Research Blog
- Society & Culture · Philosophy
- Technology
Free on the App Store for iPhone, iPad and Mac. No subscription needed. Once it is installed, tap this link again and the show is yours. Already have it? Open in Parasocial.
About
Narrations of Redwood Research blog posts. Redwood Research is a research nonprofit based in Berkeley. We investigate risks posed by the development of powerful artificial intelligence and techniques for mitigating those risks.
Also on
Recent episodes
- “A basic case for accelerating conceptual reasoning capabilities” by Emery Cooper, Caspar Oesterheld, Chi Nguyen
- “Simulated Control Evaluations” by Jeffrey Olmo, Tyler
- “Red-Teaming OpenAI’s Internal Agent Monitoring Systems” by Buck Shlegeris
- “Screwtape: Adversarial Transcript Editing for Monitor Red-Teaming” by Gosha Kozhevnikov, Tyler
- ″[Paper] Distillation for Incrimination and Distillation for Capabilities” by Sebastian Prasanna, Alek Westover
- “Frontier models’ decision theory depends on who’s asking” by Alex Kastner
- “Capabilities research expands the safety-usefulness Pareto frontier too” by Alex Mallen
- “Continual learning might make your blocking monitors nearly useless” by Alex Mallen
- “Astra is much better at reasoning with filler tokens than previous models” by Dylan Xu, Sebastian Prasanna, Alek Westover
- “Latent reasoning architectures would undermine CoT, our strongest oversight tool” by Lukas Finnveden, Alexa Pan, Alek Westover, Girish Gupta, Nathan Sheffield, Ryan Greenblatt