ACR: Adaptive Context Refactoring via Context Refactoring Operators for Multi-Turn Dialogue

Large Language Models (LLMs) have shown remarkable performance in multi-turn dialogue. However, in multi-turn dialogue, models still struggle to stay aligned with what has been established earlier, follow dependencies across many turns, and avoid drifting into incorrect facts as the interaction grows longer. Existing approaches primarily focus on extending the context window, introducing external memory, or applying context compression, yet these methods still face limitations such as $\mathbf{\text{contextual}}\mathbf{\text{inertia}}$ and $\mathbf{\text{state}}\mathbf{\text{drift}}$. To address these challenges, we propose the $\mathbf{\text{A}}$daptive $\mathbf{\text{C}}$ontext $\mathbf{\text{R}}$efactoring $\mathbf{\text{(ACR)}}$ Framework, which dynamically monitors and reshapes the interaction history to mitigate contextual inertia and state drift actively. ACR is built on a library of context refactoring operators and a teacher-guided self-evolving training paradigm that learns when to intervene and how to refactor, thereby decoupling context management from the reasoning process. Extensive experiments on multi-turn dialogue demonstrate that our method significantly outperforms existing baselines while reducing token consumption.