Rethinking the Role of Tool Graph Memory in Multi-Turn Tool Use

arXiv:2610.21571v1 Announce Type: cross Abstract: Large language model (LLM) agents solve complex tasks by calling external tools over multiple turns. Recent work has proposed tool graph memory, which stores past experience as tool transitions to guide subsequent selection. However, it remains unclear whether this component consistently improves agents' tool use or undermines it. We evaluate tool graph memory on nine models and two benchmarks, finding that it seldom improves significantly over baselines, and that its gains, where present, cost over twice the LLM calls. The memory's guidance is often incorrect, and random or no guidance retains most of the gain, implying that the benefit stems from extra computation rather than the memory itself. Moreover, even with the correct next tool suggested, many failures remain, such as giving up without a tool call. Storing the same experience instead as textual insights outperforms tool graph memory with almost no extra LLM calls. This suggests that memory serves better as a soft prior than as a prescription of the next tool, leading to a more effective and generalizable solution.

Related

• Tools Are Not Islands: Set-Level Tool Retrieval for LLM Agents via Query-Conditioned Hyperedge Prediction
• HyperAgent: Planning and Acting over Tool-Schema Hypergraphs for Tool-Use LLM Agents
• UpliftMem: Learning Set-Level Uplift for Agent Memory Retrieval
• MemToolAgent overview with a simple restaurant booking scenario where the agent retrieves similar memories, receives feedback on an invalid time format, and generates a reflection to update its memory

Source: arXiv cs.AI | 2026-10-16