Search papers, labs, and topics across Lattice.
To eliminate the redundancy, contradiction, and context-pollution common in long-horizon agent memory, the authors introduce ROAM, a framework that decouples semantic relation mapping from memory retrieval. The system classifies incoming and stored memory atom pairs across four distinct relations鈥攊ndependent, equivalent, subsuming, or conflicting鈥攁nd partitions them into active Primary views and supporting Evidence roles prior to compact fusion. Across multiple evaluation settings, this structured organization improves downstream question-answering accuracy by up to 29.8 percentage points while reducing confounder tokens by 11.5 points.
Agent memory breaks when models are forced to simultaneously interpret, update, and rewrite facts鈥攄ecoupling maintenance into explicit pairwise relation classification and role-based fusion yields up to a 29.8 percentage point accuracy leap.
Long-term language-model agents rely on external memory across interactions. Atomic memories are particularly useful: their fine-grained semantic boundaries enable precise retrieval and direct comparison between observations. Yet accumulating atoms inevitably become redundant, overlapping, or conflicting. Existing methods often ask an LLM manager to add, update, delete, or rewrite memories directly, coupling semantic interpretation, storage decisions, and content generation in one error-prone operation. We introduce ROAM, a relation-guided framework that uses atomicity for management while allowing richer answer-time representations. ROAM classifies incoming--stored atom pairs as independent, equivalent, directionally subsuming, or conflicting, then organizes observations into active Primary and supporting Evidence roles. Fusion subsequently combines complementary details and temporal changes into compact, potentially non-atomic views. Only Primary views are retrieved for answering, preventing redundant or outdated atoms from competing independently. Across models and evaluation settings, ROAM improves answer accuracy by up to 29.8 percentage points. Ablations show complementary benefits from different relations and consistent gains from fusion beyond role organization. Mechanism analysis further finds 15.6-point higher answer-critical source recall and an 11.5-point lower confounder-token share. ROAM remains robust across manager scales.
Tsinghua AI