Posts

Showing posts with the label reliability

When Does LLM Self-Correction Actually Help? Papers on Iterative Refinement, Evaluation, and Reliability