DimMem: Dimensional Structuring for Efficient Long-Term Agent Memory

arXiv cs.CL Papers

Summary

DimMem introduces a dimensional memory framework for LLM agents that represents memories as atomic, typed units with explicit fields, achieving state-of-the-art accuracy on LoCoMo-10 and LongMemEval-S while reducing token costs by 24%.

arXiv:2605.15759v1 Announce Type: new Abstract: Large language model (LLM) agents require long-term memory to leverage information from past interactions. However, existing memory systems often face a fidelity--efficiency trade-off: raw dialogue histories are expensive, while flat facts or summaries may discard the structure needed for precise recall. We propose \textbf{DimMem}, a lightweight dimensional memory framework that represents each memory as an atomic, typed, and self-contained unit with explicit fields such as time, location, reason, purpose, and keywords. This representation exposes the structure needed for dimension-aware retrieval, memory update, and selective assistant-context recall without storing full histories in the model context. Across LoCoMo-10 and LongMemEval-S, DimMem achieves \textbf{81.43\%} and \textbf{78.20\%} overall accuracy, respectively, outperforming existing lightweight memory systems while reducing LoCoMo per-query token cost by \textbf{24\%}. We further show that dimensional memory extraction is learnable by compact models: after fine-tuning on the DimMem schema, a Qwen3-4B extractor surpasses LightMem with GPT-4.1-mini on both benchmarks and reaches performance comparable to, or better than, much larger extractors in key settings. These results suggest that explicit dimensional structuring is an effective and efficient foundation for long-term memory in LLM agents. Code is available at https://github.com/ChowRunFa/DimMem.
Original Article
View Cached Full Text

Cached at: 05/18/26, 06:34 AM

# DimMem: Dimensional Structuring for Efficient Long-Term Agent Memory
Source: [https://arxiv.org/abs/2605.15759](https://arxiv.org/abs/2605.15759)
[View PDF](https://arxiv.org/pdf/2605.15759)

> Abstract:Large language model \(LLM\) agents require long\-term memory to leverage information from past interactions\. However, existing memory systems often face a fidelity\-\-efficiency trade\-off: raw dialogue histories are expensive, while flat facts or summaries may discard the structure needed for precise recall\. We propose \\textbf\{DimMem\}, a lightweight dimensional memory framework that represents each memory as an atomic, typed, and self\-contained unit with explicit fields such as time, location, reason, purpose, and keywords\. This representation exposes the structure needed for dimension\-aware retrieval, memory update, and selective assistant\-context recall without storing full histories in the model context\. Across LoCoMo\-10 and LongMemEval\-S, DimMem achieves \\textbf\{81\.43\\%\} and \\textbf\{78\.20\\%\} overall accuracy, respectively, outperforming existing lightweight memory systems while reducing LoCoMo per\-query token cost by \\textbf\{24\\%\}\. We further show that dimensional memory extraction is learnable by compact models: after fine\-tuning on the DimMem schema, a Qwen3\-4B extractor surpasses LightMem with GPT\-4\.1\-mini on both benchmarks and reaches performance comparable to, or better than, much larger extractors in key settings\. These results suggest that explicit dimensional structuring is an effective and efficient foundation for long\-term memory in LLM agents\. Code is available at[this https URL](https://github.com/ChowRunFa/DimMem)\.

## Submission history

From: Wentao Qiu \[[view email](https://arxiv.org/show-email/1092977b/2605.15759)\] **\[v1\]**Fri, 15 May 2026 09:20:31 UTC \(1,172 KB\)

Similar Articles

HeLa-Mem: Hebbian Learning and Associative Memory for LLM Agents

arXiv cs.CL

HeLa-Mem is a bio-inspired memory architecture for LLM agents that models memory as a dynamic graph using Hebbian learning dynamics, featuring episodic and semantic memory stores to improve long-term coherence. Experiments on LoCoMo show superior performance across question categories while using fewer context tokens.