English

Is One Score Enough? Rethinking the Evaluation of Sequentially Evolving LLM Memory

Machine Learning 2026-05-18 v1 Artificial Intelligence

Abstract

Memory plays a central role in enabling large language models (LLMs) to operate over sequential tasks by accumulating and reusing experience over time. However, existing evaluations of LLM memory mostly rely on aggregate metrics such as final hold-out accuracy or cumulative online performance, which can obscure critical failure modes such as forgetting and negative transfer. In this paper, we introduce SeqMem-Eval, a diagnostic evaluation framework for sequentially evolving LLM memory. Drawing inspiration from continual learning, it targets a test-time setting in which memory is external, prompt-mediated, and updated without modifying model parameters. Rather than focusing only on final performance, SeqMem-Eval evaluates how memory states evolve, generalize, consolidate experience, and retain useful information during sequential inference. Specifically, it measures online utility, hold-out generalization, backward transfer, and forgetting, providing a finer-grained view of memory quality. Through extensive experiments across diverse tasks and memory methods, we show that higher final or cumulative accuracy does not necessarily imply better memory quality: many methods exhibit strong performance gains while suffering from substantial forgetting or negative transfer. Moreover, different memory designs exhibit distinct trade-offs between adaptability and stability that remain invisible under standard evaluation metrics.

Keywords

Cite

@article{arxiv.2605.15384,
  title  = {Is One Score Enough? Rethinking the Evaluation of Sequentially Evolving LLM Memory},
  author = {Songwei Dong and Zihan Chen and Chengshuai Shi and Peng Wang and Jundong Li and Cong Shen},
  journal= {arXiv preprint arXiv:2605.15384},
  year   = {2026}
}

Comments

29 pages, 13 figures

R2 v1 2026-07-22T07:13:19.083Z