MEGA Hub

REMORY: Learning Residual Memory for Context Compaction

Authors

Do you know Hanchen Xia?You can claim authorship or link another user.Do you know Baoyou Chen?You can claim authorship or link another user.Do you know Yutang Ge?You can claim authorship or link another user.Do you know Naihao Deng?You can claim authorship or link another user.Do you know Senqiao Yang?You can claim authorship or link another user.Do you know Zilong Dong?You can claim authorship or link another user.Do you know Weihao Yuan?You can claim authorship or link another user.Do you know Siyu Zhu?You can claim authorship or link another user.

Abstract

Long-horizon agents compact their history to continue within a finite context window, but a textual summary alone may not support every subsequent decision. We introduce REMORY, a neural memory network that supplements the summary with a bounded sequence of soft memory tokens. Given the history and summary, the network learns to generate tokens that help a frozen LLM approximate the continuation it would produce with the full history. The tokens are conditioned on the summary and appended after it, forming an analogue of a residual connection along the sequence dimension. On SummHay, REMORY improves source attribution at nearly unchanged insight coverage and approaches the full-context joint score using only 5.2% of the input positions. Across long-horizon agent benchmarks, Qwen3.8-27B and GLM-5.3-Flash show consistent gains with residual memory. Both models also exhibit substantially fewer repeated tool outputs and tool errors on BrowseComp and Terminal-Bench 2.1.

Community

00