記憶是重構而非提取：大型語言模型代理的圖形記憶

摘要

儘管近期有所進展，大型語言模型代理在處理長篇互動歷程的推理時仍面臨困難。現行的記憶增強代理依賴於靜態的「先擷取再推理」典範，這種僵化的管線設計使得它們無法根據推理過程中發現的中間證據動態調整記憶存取方式。為了解決這個問題，我們提出了MRAgent框架，該框架結合了關聯記憶圖與主動重構機制。我們將記憶表示為「線索-標籤-內容」圖，其中關聯標籤作為語義橋樑，連接細粒度線索與記憶內容。在此結構上運作的主動重構機制，將大型語言模型的推理直接整合至記憶存取中，使代理能根據累積的證據逐步探索並剪裁檢索路徑。此舉確保記憶檢索能根據推理脈絡動態調整，同時避免因無限制擴張而導致的組合爆炸問題。在LoCoMo基準與LongMemEval基準上的實驗結果顯示，相較於強基線方法，我們的方法在效能上顯著提升（最高達23%），同時大幅降低詞元消耗與執行時間成本，凸顯了主動與關聯重構在長時程記憶推理中的有效性。

English

Despite recent progress, LLM agents still struggle with reasoning over long interaction histories. While current memory-augmented agents rely on a static retrieve-then-reason paradigm, this rigid pipeline design prevents them from dynamically adapting memory access to intermediate evidence discovered during inference. To bridge this gap, we propose MRAgent, a framework that combines an associative memory graph with an active reconstruction mechanism. We represent memory as a Cue-Tag-Content graph, where associative tags serve as semantic bridges connecting fine-grained cues to memory contents. Operating on this structure, our active reconstruction mechanism integrates LLM reasoning directly into memory access, allowing the agent to iteratively explore and prune retrieval paths based on accumulated evidence. This ensures that memory retrieval is dynamically adapted to the reasoning context while avoiding combinatorial explosion caused by unconstrained expansion. Experiments on the LoCoMo benchmark and LongMemEval benchmark demonstrate significant improvements over strong baselines (up to 23%), while substantially reducing token and runtime cost, highlighting the effectiveness of active and associative reconstruction for long-horizon memory reasoning.