# LAM: Efficient Lossy Agent Memory Framework With A Retrieval-Score Error Bound

arXiv: https://arxiv.org/abs/2609.32256 · HTML: https://mlsanity.com/paper/2609.32256/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Agent memory systems, one of 16 directions in LLM agents and tool use.
- Builds on 24 of its 36 references that are in the map.
- First posted on arXiv 26 Sep 2026.

## Placed in

- LLM agents and tool use / [Agent memory systems](https://mlsanity.com/agents/agent-memory-systems/)

## Builds on (most cited first, 24 of 24)

- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory](https://mlsanity.com/paper/2504.19413/) (2025)
- [Learning to Compress Prompts with Gist Tokens](https://mlsanity.com/paper/2304.08467/) (2023)
- [SemDeDup: Data-efficient learning at web-scale through semantic deduplication](https://mlsanity.com/paper/2303.09540/) (2023)
- [MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent](https://mlsanity.com/paper/2507.02259/) (2025)
- [In-context Autoencoder for Context Compression in a Large Language Model](https://mlsanity.com/paper/2307.06945/) (2023)
- [LightMem: Lightweight and Efficient Memory-Augmented Generation](https://mlsanity.com/paper/2510.18866/) (2025)
- [Cost-Efficient Large Language Model Serving for Multi-turn Conversations with CachedAttention](https://mlsanity.com/paper/2403.19708/) (2024)
- [xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Token](https://mlsanity.com/paper/2405.13792/) (2024)
- [Vidur: A Large-Scale Simulation Framework For LLM Inference](https://mlsanity.com/paper/2405.05465/) (2024)
- [Preble: Efficient Distributed Prompt Scheduling for LLM Serving](https://mlsanity.com/paper/2407.00023/) (2024)
- [The Complexity Trap: Simple Observation Masking Is as Efficient as LLM Summarization for Agent Context Management](https://mlsanity.com/paper/2508.21433/) (2025)
- [Fundamental Limits of Prompt Compression: A Rate-Distortion Framework for Black-Box Language Models](https://mlsanity.com/paper/2407.15504/) (2024)
- [Prompt Compression in the Wild: Measuring Latency, Rate Adherence, and Quality for Faster LLM Inference](https://mlsanity.com/paper/2604.02985/) (2026)
- [An Empirical Study on Prompt Compression for Large Language Models](https://mlsanity.com/paper/2505.00019/) (2025)
- [Remember the Decision, Not the Description: A Rate-Distortion Framework for Agent Memory](https://mlsanity.com/paper/2605.10870/) (2026)
- [KVServe: Service-Aware KV Cache Compression for Communication-Efficient Disaggregated LLM Serving](https://mlsanity.com/paper/2605.13734/) (2026)
- [Agent Memory: Characterization and System Implications of Stateful Long-Horizon Workloads](https://mlsanity.com/paper/2606.06448/) (2026)
- [Slipstream: Trajectory-Grounded Compaction Validation for Long-Horizon Agents](https://mlsanity.com/paper/2605.08580/) (2026)
- [Parallel Context Compaction for Long-Horizon LLM Agent Serving](https://mlsanity.com/paper/2605.23296/) (2026)
- [Control Under Compression: Reliability Frontiers for Tool-Using Agents](https://mlsanity.com/paper/2608.01056/) (2026)
- [SMetric: Rethink LLM Scheduling for Serving Agents with Balanced Session-centric Scheduling](https://mlsanity.com/paper/2607.08565/) (2026)
- [Prompt Compression in Production Task Orchestration: A Pre-Registered Randomized Trial](https://mlsanity.com/paper/2603.23525/) (2026)
- [KVMem: Virtualizing Million-Token Agent Workspaces on a Consumer GPU](https://mlsanity.com/paper/2609.04852/) (2026)

## Abstract (arXiv)

Agent memory grows as agents read inputs, reason, and call tools. Longer histories increase inference cost and eventually exceed the context window. LLM-based summarization reduces this history but adds latency and provides no explicit bound on information loss. We propose LAM, a Lossy Agent Memory system with three components: a deterministic deduplication rule with a substitution bound on retrieval scores - a bound on score perturbation, not a certificate of unchanged ranking; a memory manager that preserves the cached prefix and overlaps compaction with inference; and a performance model that estimates compaction costs before deployment. On 600 agent trajectories, LAM removes 22.47% of observation tokens while retaining 99.984% of the measured gold-patch evidence. At a fixed deletion set, the performance model predicts a 71.4x-91.6x end-to-end speedup from removing records before prefill instead of deleting them from a prefilled context. That benefit comes from the schedule rather than the rule and applies to any prefix-preserving test.