记忆-T1：多会话智能体中时序推理的强化学习方法

摘要

针对长程多轮对话的时间推理能力是对话智能体的核心需求。然而现有研究及我们的初步实验表明，当对话历史增长并累积噪声时，当前的长上下文模型难以准确识别时间相关信息，严重影响了推理性能。为此，我们提出Memory-T1框架，该框架通过强化学习训练具有时间感知能力的记忆选择策略。其采用由粗到精的双阶段处理：首先通过时间与相关性过滤器对对话历史进行剪枝生成候选集，再由强化学习智能体精准筛选证据会话。强化学习训练受多层级奖励函数引导，同步优化（i）答案准确性、（ii）证据可追溯性及（iii）时间一致性。其中时间一致性奖励通过评估会话级（时序邻近性）和语句级（时序保真度）与查询时间范围的匹配度，提供密集信号，使智能体能够解析细微的时间歧义。在Time-Dialog基准测试中，Memory-T1将70亿参数模型的综合得分提升至67.0%，创造了开源模型的新标杆，较140亿参数基线模型提升10.2%。消融实验表明时间一致性与证据可追溯性奖励共同带来15.0%的性能增益。此外，当基线模型在12.8万令牌规模下失效时，Memory-T1仍保持稳健性能，证明了其对长对话历史噪声的有效抑制。代码与数据集已开源：https://github.com/Elvin-Yiming-Du/Memory-T1/

English

Temporal reasoning over long, multi-session dialogues is a critical capability for conversational agents. However, existing works and our pilot study have shown that as dialogue histories grow in length and accumulate noise, current long-context models struggle to accurately identify temporally pertinent information, significantly impairing reasoning performance. To address this, we introduce Memory-T1, a framework that learns a time-aware memory selection policy using reinforcement learning (RL). It employs a coarse-to-fine strategy, first pruning the dialogue history into a candidate set using temporal and relevance filters, followed by an RL agent that selects the precise evidence sessions. The RL training is guided by a multi-level reward function optimizing (i) answer accuracy, (ii) evidence grounding, and (iii) temporal consistency. In particular, the temporal consistency reward provides a dense signal by evaluating alignment with the query time scope at both the session-level (chronological proximity) and the utterance-level (chronological fidelity), enabling the agent to resolve subtle chronological ambiguities. On the Time-Dialog benchmark, Memory-T1 boosts a 7B model to an overall score of 67.0\%, establishing a new state-of-the-art performance for open-source models and outperforming a 14B baseline by 10.2\%. Ablation studies show temporal consistency and evidence grounding rewards jointly contribute to a 15.0\% performance gain. Moreover, Memory-T1 maintains robustness up to 128k tokens, where baseline models collapse, proving effectiveness against noise in extensive dialogue histories. The code and datasets are publicly available at https://github.com/Elvin-Yiming-Du/Memory-T1/

记忆-T1：多会话智能体中时序推理的强化学习方法

Memory-T1: Reinforcement Learning for Temporal Reasoning in Multi-session Agents

摘要

Support