Redwood ResearchResearch
Continual learning might make your blocking monitors nearly useless
When monitor evasion looks like legitimate learning to your continual learning system, it’s hard to have one without the other
Read at Redwood Research ↗More from Redwood Research on Accept All
ResearchLatent reasoning architectures would likely undermine CoT, our strongest oversight tool Redwood Research
ResearchAstra is much better at reasoning with filler tokens than previous models Redwood Research
ResearchCoT controllability evals seem very under-elicited Redwood Research
ResearchAn operationalization of opaque serial depth Redwood Research
ResearchProposal for tracking the effects of architecture on monitorability Redwood Research
