面向 LLM 多轮交互的测试时策略自适应方法 T2PAM 与 ROSA 算法

View PDF HTML (experimental)

Abstract:Large Language Models (LLMs) employ multi-turn interaction as a fundamental paradigm for completing complex tasks. However, their performance often degrades in extended interactions, as they are typically trained on static, single-turn data, which hinders their ability to adapt to real-time user feedback. To address this limitation, we first propose a new paradigm: Test-Time Policy Adaptation for Multi-Turn Interactions (T2PAM), which utilizes user feedback from the ongoing interaction as a reward signal to estimate a latent optimal policy aligned with user preferences, then updates a small subset of parameters to steer the model toward this policy, ultimately enabling efficient in-conversation self-correction. We then introduce Optimum-Referenced One-Step Adaptation (ROSA), a lightweight algorithm that operationalizes T2PAM. ROSA guides the model parameters toward a theoretical optimal policy in a single, efficient update step, avoiding costly iterative gradient-based optimization and minimizing computational overhead. We provide a rigorous theoretical analysis guaranteeing that the policy of ROSA converges to the preference of user as the number of interactions increases. Extensive experiments on challenging benchmark demonstrate that ROSA achieves significant improvements in both task effectiveness and efficiency.

Submission history

From: Chenxing Wei [view email]
[v1] Sat, 27 Sep 2025 07:46:15 UTC (743 KB)
[v2] Mon, 2 Mar 2026 02:32:21 UTC (728 KB)
[v3] Sun, 27 Sep 2026 05:19:50 UTC (777 KB)
[v4] Mon, 26 Oct 2026 05:19:47 UTC (777 KB)