From Volume to Value: Preference-Aligned Memory Construction for On-Device RAG
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914577888837632 |
|---|---|
| author | Lee, Changmin Kim, Jaemin Gong, Taesik |
| author_facet | Lee, Changmin Kim, Jaemin Gong, Taesik |
| contents | With the rapid emergence of personal AI agents based on Large Language Models (LLMs), implementing them on-device has become essential for privacy and responsiveness. To handle the inherently personal and context-dependent nature of real-world requests, such agents must ground their generation in device-resident personal context. However, under tight memory budgets, the core bottleneck is what to store so that retrieval remains aligned with the user. We propose EPIC (Efficient Preference-aligned Index Construction), which focuses on user preferences as a compact and stable form of personal context and integrates them throughout the RAG pipeline. EPIC selectively retains preference-relevant information from raw data and aligns retrieval toward preference-aligned contexts. Across four benchmarks covering conversations, debates, explanations, and recommendations, EPIC reduces indexing memory by 2,404 times, improves preference-following accuracy by 20.17 percentage points, and achieves 33.33 times lower retrieval latency over the best-performing baseline. In our on-device experiment, EPIC maintains a memory footprint under 1 MB with 29.35 ms/query latency in streaming updates. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_18271 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | From Volume to Value: Preference-Aligned Memory Construction for On-Device RAG Lee, Changmin Kim, Jaemin Gong, Taesik Computation and Language Artificial Intelligence Information Retrieval Machine Learning With the rapid emergence of personal AI agents based on Large Language Models (LLMs), implementing them on-device has become essential for privacy and responsiveness. To handle the inherently personal and context-dependent nature of real-world requests, such agents must ground their generation in device-resident personal context. However, under tight memory budgets, the core bottleneck is what to store so that retrieval remains aligned with the user. We propose EPIC (Efficient Preference-aligned Index Construction), which focuses on user preferences as a compact and stable form of personal context and integrates them throughout the RAG pipeline. EPIC selectively retains preference-relevant information from raw data and aligns retrieval toward preference-aligned contexts. Across four benchmarks covering conversations, debates, explanations, and recommendations, EPIC reduces indexing memory by 2,404 times, improves preference-following accuracy by 20.17 percentage points, and achieves 33.33 times lower retrieval latency over the best-performing baseline. In our on-device experiment, EPIC maintains a memory footprint under 1 MB with 29.35 ms/query latency in streaming updates. |
| title | From Volume to Value: Preference-Aligned Memory Construction for On-Device RAG |
| topic | Computation and Language Artificial Intelligence Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2605.18271 |