Preprint · 2026

When Does Memory Help? A Cost-Aware Evaluation of Long-Term Memory in Tool-Using LLM Agents

Shweta Mishra, Shashank Mishra

Abstract

Long-term memory for LLM agents is evaluated today by conversational recall benchmarks (LoCoMo, LongMemEval), which measure question answering over dialogue history, not whether remembered facts change what a tool-using agent does. We present MERIT (Memory Evaluation for Realistic Instrumented Tasks), a benchmark and harness that measures the marginal utility of memory for task-executing agents under explicit cost accounting. MERIT provides episodic tool-use tasks in three domains whose dependence on earlier-episode facts is verified by an automated leak check; a difficulty ladder ending in updated-fact recall; controlled memory corruption; and full token and dollar metering of every memory operation. Across 23,440 scored episodes ($42.57), a two-generation pilot on gpt-4.1-mini and a preregistered 3-model x 3-seed grid (GPT-4.1, Claude Haiku 4.5; memory side held fixed), memory lifts dependent-task success from a leak-verified floor of 0.00 to 0.55-1.00. On updated facts, embedding retrieval collapses unpredictably (0.30-0.95 across models; max seed gap 0.45), and agents act on a correctly retrieved value only 55% of the time, while update-on-write stores (a structured fact store and, notably, LLM summarization) remain at 0.70-1.00; the hybrid is worse than the fact store alone. A latest-generation spot-check (Claude Sonnet 5, gated on a clean full-replay control) reproduces the pattern. Swapping a memory's implementation moves task success by up to 60 points, and full replay is never economical: the best condition per domain delivers 2.7-3.9x its marginal utility per dollar. We release the benchmark, harness, and all traces.

BibTeX Copy

@misc{mishra_when_2026,
	title = {When {Does} {Memory} {Help}? {A} {Cost}-{Aware} {Evaluation} of {Long}-{Term} {Memory} in {Tool}-{Using} {LLM} {Agents}},
	shorttitle = {When {Does} {Memory} {Help}?},
	url = {http://arxiv.org/abs/2609.05441},
	doi = {10.48550/arXiv.2609.05441},
	abstract = {Long-term memory for LLM agents is evaluated today by conversational recall benchmarks (LoCoMo, LongMemEval), which measure question answering over dialogue history, not whether remembered facts change what a tool-using agent does. We present MERIT (Memory Evaluation for Realistic Instrumented Tasks), a benchmark and harness that measures the marginal utility of memory for task-executing agents under explicit cost accounting. MERIT provides episodic tool-use tasks in three domains whose dependence on earlier-episode facts is verified by an automated leak check; a difficulty ladder ending in updated-fact recall; controlled memory corruption; and full token and dollar metering of every memory operation. Across 23,440 scored episodes (\$42.57), a two-generation pilot on gpt-4.1-mini and a preregistered 3-model x 3-seed grid (GPT-4.1, Claude Haiku 4.5; memory side held fixed), memory lifts dependent-task success from a leak-verified floor of 0.00 to 0.55-1.00. On updated facts, embedding retrieval collapses unpredictably (0.30-0.95 across models; max seed gap 0.45), and agents act on a correctly retrieved value only 55\% of the time, while update-on-write stores (a structured fact store and, notably, LLM summarization) remain at 0.70-1.00; the hybrid is worse than the fact store alone. A latest-generation spot-check (Claude Sonnet 5, gated on a clean full-replay control) reproduces the pattern. Swapping a memory's implementation moves task success by up to 60 points, and full replay is never economical: the best condition per domain delivers 2.7-3.9x its marginal utility per dollar. We release the benchmark, harness, and all traces.},
	urldate = {2026-11-06},
	publisher = {arXiv},
	author = {Mishra, Shweta and Mishra, Shashank},
	month = jul,
	year = {2026},
	note = {arXiv:2609.05441 [cs.AI]},
	keywords = {Computer Science - Artificial Intelligence},
}