cs.CLOct 30, 2026

KEEPER: Learning Control-Sufficient Textual Memory for Long-Horizon LLM Agents

Authors: Jinkyeong Choi, Chaebin Jeong, Junsung Jeon, Donghyeon Park

Organizations: Sejong University, Seoul, Republic of Korea

Abstract

Recursive textual memory can carry earlier errors into later updates, yet task performance alone does not establish memory reliability. We define self-consistency bias as the persistence of incorrect claims already present in the incoming memory. We present KEEPER, a comparative analysis of four auxiliary objectives along information-content and update-sensitivity axes within a shared recursive interface, examining how supervision targets relate to agent performance. Among evaluated conditions, reward-related prediction achieve the highest task performance before and after offline reinforcement learning, outperforming joint supervision. A claim-level analysis of one baseline trajectory documents error persistence despite corrective observations. Together, these analyses motivate evaluating error persistence and valid-information retention alongside task outcomes.

Figures & tables

Appendix figures & tables3 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. MemTrain: Self-Supervised Context Memory Training

    Jun 2, 2026Ziheng Li, Xingrun Xing, Haoqing Wang +2LLM Agent MemoryPersistent Memory for Language Models

  2. From Reflection to Co-Evolution: Self-Evolving Agents with Locally Verified Memory

    Nov 7, 2026Xinxiang Yin, Xinglang Zhang, Haodong Chen +7Memory Management for LLM AgentsExperience Learning for LLM Agents