ConceptGenerator-Selector Framework
Generator-Selector Framework
An agent reasoning framework in which a generator produces a diverse set of candidate actions and a selector acts as a scoring function that chooses the best one for execution. The selector may be implemented as a reward model, a critic model, or a general-purpose LLM.
Timeline (last 90 days)
Nov 5
Turn-level Best-of-N sampling is an inference-time scaling method that improves the performance of LLM agents on long-horizon tasks through a generator-selector framework.
Unverified50%