Beyond the Black Box: Demystifying Multi-Turn LLM Reasoning with VISTA

Yiran Zhang, Mingyang Lin, Mark Dras +1

Recent research has increasingly focused on the reasoning capabilities of Large Language Models (LLMs) in multi-turn interactions, as these scenarios more closely mirror real-world…