Direct-Inverse Prompting: Analyzing LLMs' Discriminative Capacity in Self-Improving Generation

Ryo Kamoi

Ryo Kamoi is a PhD candidate at Penn State University, supported by 2026 Google PhD Fellowship. His research focuses on self-evolving LLM agents and LLM-based verification of LLM responses and intermediate steps.