Mem2ActBench: A Benchmark for Evaluating Long-Term Memory Utilization in Task-Oriented Autonomous Agents

TL;DR

Mem2ActBench is a new benchmark designed to evaluate whether autonomous agents can proactively use long-term memory to select tools and execute tasks effectively, addressing limitations of existing passive retrieval benchmarks.

Contribution

The paper introduces Mem2ActBench, a comprehensive benchmark with a new dataset and evaluation method for assessing active long-term memory utilization in task-oriented agents.

Findings

Current systems are inadequate at actively leveraging memory for task execution.

The benchmark reveals significant gaps in existing memory frameworks.

Human evaluation confirms high memory dependency in generated tasks.

Abstract

Large Language Model (LLM)-based agents are increasingly deployed for complex, tool-based tasks where long-term memory is critical to driving actions. Existing benchmarks, however, primarily test a angent's ability to passively retrieve isolated facts in response to explicit questions. They fail to evaluate the more crucial capability of actively applying memory to execute tasks. To address this gap, we introduce \textsc{Mem2ActBench}, a benchmark for evaluating whether agents can proactively leverage long-term memory to execute tool-based actions by selecting appropriate tools and grounding their parameters. The benchmark simulates persistent assistant usage, where users mention the same topic across long, interrupted interactions and expect previously established preferences and task states to be implicitly applied. We build the dataset with an automated pipeline that merges heterogeneous sources (ToolACE, BFCL, Oasst1), resolves conflicts via consistency modeling, and synthesizes 2,029 sessions with 12 user--assistant--tool turns on average. From these memory chains, a reverse-generation method produces 400 tool-use tasks, with human evaluation confirming 91.3\% are strongly memory-dependent. Experiments on seven memory frameworks show that current systems remain inadequate at actively utilizing memory for parameter grounding, highlighting the need for more effective approaches to evaluate and improve memory application in task execution.