# MemRouter: Memory-as-Embedding Routing for Long-Term Conversational Agents

Source: [MemRouter: Memory-as-Embedding Routing for Long-Term Conversational Agents](<https://www.sota2.com/research/paper/69f82813c0f56ed843088a95>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-05T05:00:04.460Z

Authors: Tianyu Hu, Weikai Lin, Weizhi Zhang, Jing Ma, Song Wang

Published: 2026-05-01T02:33:50.000Z

## Abstract

Long-term conversational agents must decide which turns to store in external memory, yet recent systems rely on autoregressive LLM generation at every turn to make that decision. We present MemRouter, a write-side memory router that decouples memory admission from the downstream answer backbone and replaces per-turn memory-management decoding with an embedding-based routing policy. MemRouter encodes each turn together with recent context, projects the resulting embeddings through a frozen LLM backbone, and predicts whether the turn should be stored using lightweight classification heads while training only 12M parameters. Under a controlled matched-harness comparison on LoCoMo, where the retrieval pipeline, answer prompts, and QA backbone (Qwen2.5-7B) are held identical, MemRouter outperforms an LLM-based memory manager on every question category (overall F1 52.0 vs 45.6, non-overlapping 95% CIs) while reducing memory-management p50 latency from 970ms to 58ms. Descriptive factorial averaging further shows that learned admission improves mean F1 by +10.3 over random storage, category-specific prompting adds +5.2 over a generic prompt, and retrieval contributes +0.7. These results suggest that write-side memory admission can be learned by a small supervised router, while answer generation remains a separate downstream component in long-horizon conversational QA.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Question Answering on LoCoMo excluding adversarial (test)](<https://www.sota2.com/research/sota/question-answering-on-locomo-excluding-adversarial-test.md>): Question Answering; LoCoMo excluding adversarial (test); Overall F1: 55.5.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2605.00356>)

[PDF](<https://arxiv.org/pdf/2605.00356.pdf>)