Episode
"Proposal for tracking the effects of architecture on monitorability" by ryan_greenblatt, Alek Westover, Lukas Finnveden
- Published
- Sep 11, 2026
- Duration seconds
- 676
- Processing state
not_requested
Actions
POST https://stenobird.com/v1/public/podcasts/lesswrong-curated-popular-5643401/episodes/proposal-for-tracking-the-effects-of-architecture-on-monitorability-by-ryan-greenblatt-alek-westover-lukas-finnveden/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/lesswrong-curated-popular-5643401/proposal-for-tracking-the-effects-of-architecture-on-monitorability-by-ryan-greenblatt-alek-westover-lukas-finnveden.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
Architectures that incorporate opaque recurrence or allow for agents to communicate with each other using latents could rapidly make it much harder to monitor chains of thought or communication (we’ll refer to this property as “monitorability” going forward). As companies begin to explore such architectures, we believe it is important to transparently share evidence about how monitorability varies with architecture and training method. To inform the scientific debate on how to make tradeoffs...