LGAIJul 18, 2025

A Simple "Try Again" Can Elicit Multi-Turn LLM Reasoning

arXiv:2507.14295v211 citationsh-index: 8Has Code
Originality Incremental advance
AI Analysis

This addresses the problem of repetitive and inflexible responses in multi-turn AI reasoning for users needing iterative problem-solving, though it is incremental as it builds on existing single-turn RL methods.

The paper tackles the challenge of enabling Large Reasoning Models to effectively revise answers in multi-turn problem-solving contexts by introducing Unary Feedback as Observation (UFO), a reinforcement learning method that uses minimal feedback like 'try again' to improve multi-turn reasoning accuracy by up to 14% while maintaining single-turn performance.

Multi-turn problem solving is critical yet challenging for Large Reasoning Models (LRMs) to reflect on their reasoning and revise from feedback. Existing Reinforcement Learning (RL) methods train large reasoning models on a single-turn paradigm with verifiable rewards. However, we observe that models trained with existing RL paradigms often lose their ability to solve problems across multiple turns and struggle to revise answers based on contextual feedback, leading to repetitive responses. We ask: can LRMs learn to reflect their answers in a multi-turn context? In this work, we find that training models with multi-turn RL using only unary feedback (e.g., "Let's try again") after wrong answers can improve both single-turn performance and multi-turn reasoning. We introduce Unary Feedback as Observation (UFO) for reinforcement learning, which uses minimal yet common unary user feedback during iterative problem solving. It can be easily applied to existing single-turn RL training setups. Experimental results show that RL training with UFO keeps single-turn performance and improves multi-turn reasoning accuracy by up to 14%, enabling language models to better react to feedback in multi-turn problem solving. To further minimize the number of turns needed for a correct answer while encouraging diverse reasoning when mistakes occur, we design reward structures that guide models to produce careful and deliberate answers in each turn. Code: https://github.com/lichengliu03/unary-feedback

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes