arXiv cs.CL· Senhao Liu, Yuheng Zhang, Chunyu Wei, Yueguo Chen, Xinran Zhang·· 10 小时前AI 评分38
OpticalRAG：面向 Token 高效检索增强生成的像素空间压缩框架
OpticalRAG: Pixel-Space Compression for Token-Efficient Retrieval-Augmented Generation
AI 导读
OpticalRAG 将文档渲染为页面图像，用冻结视觉编码器编码后把检索到的视觉 token 蒸馏为仅 16 个注入 token 再送入 LLM。在 LongBench-v2、BAMBOO 和 LooGLE-v2 三个基准上，该方法均取得最佳压缩感知准确率（CAP），并使闭卷 Qwen2.5-7B-Instruct 主干模型提升 5.17 分。
正文 · 原文
Abstract:Retrieval-augmented generation (RAG) is bottlenecked by the token budget of large language models: feeding long documents into the context window is expensive, while aggressive textual compression discards fine-grained information that is irreversibly lost in a one-dimensional token sequence. We argue that this bottleneck is not fundamental to information density but to the choice of compression \emph{modality}. Modern visual encoders can pack the contents of a rendered text page into roughly a hundred visual tokens with little semantic loss, opening a new compression-fidelity tradeoff that purely textual methods cannot reach. We present \textbf{OpticalRAG}, a framework that renders documents as page images, encodes them with a frozen vision encoder, and distills the retrieved visual tokens into as few as $16$ injected tokens before passing them to the LLM. Two ingredients make this work without expensive cross-modal training: (i) \emph{Encoder-Bridged Retrieval}, which projects text-query states into the visual feature space of the frozen encoder, through a projection shared with the compressor, to retrieve visual tokens from a text query, with no contrastive training; and (ii) \emph{Query-Driven Distillation}, a lightweight cross-attention compressor that condenses the retrieved visual tokens into a small set of query-conditioned prefix tokens. On LongBench-v2, BAMBOO, and LooGLE-v2, OpticalRAG achieves the best compression-aware accuracy (CAP) on all three benchmarks while injecting only $16$ tokens, and improves the closed-book Qwen2.5-7B-Instruct backbone by $5.17$ points on this http URL is available at this https URL.
| Comments: | Accepted at NeurIPS 2026 |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.13675 [cs.CL] |
| (or arXiv:2611.13675v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.13675 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yuheng Zhang [view email]
[v1]
Tue, 10 Nov 2026 07:24:50 UTC (386 KB)
来源：arXiv cs.CL · arxiv.org