The Q* hypothesis: Tree-of-thoughts reasoning, process reward models, and supercharging synthetic data
Nathan Lambertinterconnects.ai
The Q* hypothesis: Tree-of-thoughts reasoning, process reward models, and supercharging synthetic data
The more principled path forward, Samuel reasoned, was for the computer itself to somehow generate strategic considerations on its own.
Brian Christian • The Alignment Problem
Should the Q-value contain the expected future rewards that you could earn from taking this action? Or the expected rewards that you would earn? For a totally perfect agent, there is no tension—but otherwise the prescriptions can vary sharply.
Brian Christian • The Alignment Problem
There are two different elements to the problem of Alignment. Getting an AI to do the things we want, and being able to come to terms on what we actually want. We’ve gotta align the AI to the humans, and we also gotta align the humans to the other humans (both present and future). My idea takes from my experience in how DAOs and other mechanisms tr... See more
Prometheus • Using Consensus Mechanisms as an approach to Alignment — LessWrong
Inverse reinforcement learning is, famously, what mathematicians call an “ill-posed” problem: namely, one that doesn’t have a single, unique right answer.