An operationalization of opaque serial depth

T2Covered by 1 source · first reported by AI Alignment Forum / LessWrong (curated) at 10 Sept, 17:33 UTC

Currently, chain-of-thought (CoT) is a valuable tool for overseeing AI models. However, some architectural shifts could significantly reduce CoT monitorability. We have recently proposed that AI companies should transparently share information about the degree to which their architectures may…