Using Chunked Monitoring to Detect Deception in Long Transcripts

·LessWrong··

Summary: Monitoring long transcripts in chunks, rather than all at once, catches behaviors previously missed.We pose chunked monitoring as an effective means of finding the needle in the haystack, where the “needle” is deceptive behavior. Our findings include:The chunked monitor beat the global one in the partial_problem_solving (recovered ~93% Global FN), match_weaker_model (recovered ~86% Global FN), and refusals (recovered ~45% Global FN) categories of sandbagging.The chunked monitor consiste...

Read full article →

Related Articles

AI has access to a vastly larger working memory than the human brain
rzk · Hacker News · 4h ago
Semaglutide linked to lower predicted dementia risk
randycupertino · Hacker News · 6h ago
Firefox is now the last major browser that still supports uBlock Origin
DemiGuru · Hacker News · 1d ago
GLM-5.3: Frontier coding with emergent cyber capabilities
pella · Hacker News · 1d ago
Going Dark, and the era of law enforcement hacking
vslira · Hacker News · 1d ago