DELTAMEM: Incremental Experience Memory for LLM Agents via Residual Trees
Quick Take
DeltaMem introduces a novel memory framework for LLM agents, organizing experiences into two residual trees to reduce redundancy and improve retrieval accuracy. Experiments show DeltaMem outperforms existing baselines across various interactive environments, enhancing learning efficiency.
Key Points
- DeltaMem organizes experience memory into goal-conditioned and scene-level residual trees.
- Each tree uses root nodes for base experiences and delta nodes for variations.
- A failure-penalized similarity scan enhances retrieval of relevant experiences.
- Autonomous consolidation distills frequent paths into new root nodes.
- DeltaMem consistently outperforms existing memory frameworks in experiments.
Article Content
From source RSS / original summaryarXiv:2606. 03083v1 Announce Type: new Abstract: Large Language Model (LLM)-based agents increasingly rely on memory to learn from experiences over continual interactions. However, storing experiences as independent, flat units leads to substantial redundancy and retrieval conflicts, as similar episodes repeat overlapping content and subtle scene variations cause retrieved memories to offer contradictory guidance.
To address this, we introduce residual experience, positing that newly acquired experience is often an incremental variation of existing knowledge. We propose DeltaMem, a framework that organizes experience memory into two independent residual trees, one storing goal-conditioned task experience as reusable skills and another for scene-level environment knowledge.
Each tree uses a root node for generalized base experiences and incremental delta nodes for subsequent variations, allowing related experiences to share a common foundation without duplication. For retrieval, a failure-penalized similarity scan locates the best match, reconstructing the full experience via root-to-match chain composition. An autonomous consolidation mechanism distills high-frequency paths into new root nodes, enabling the trees to self-organize from general heuristics to specialized variants.
Experiments across diverse interactive environments show that DeltaMem consistently outperforms existing baselines. To facilitate future research, we release the code at https://github. com/import-myself/DeltaMem.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →AUDITFLOW: Executable Symbolic Environments for Structured Financial Reporting Verification
AuditFlow introduces a multi-agent framework for structured financial reporting verification, achieving 82.09% accuracy with GPT-5.5, outperforming the baseline by 14.93 points. It utilizes a symbolic environment for effective audit processes, demonstrating the necessity of deterministic checks for reliable verification.