收录解读
依赖OCR文本的文档研究可能损失图表、公式及版面线索,论文探索直接以页面图像为检索和推理单位的科学文档研究流程。
DR Pixel在106万页科学文档语料上构建页面检索与视觉推理代理,并配套DRPixelBench和用于轨迹训练的反事实信用重加权方法。
106万页面的页图检索研究流程与反事实步骤加权形成可复用方法;训练/测试支撑论文分离,三次运行;自建集准确率40.1→52.1,外部三科学集提高9.6–14.4个百分点,ViDoRe跨域提高4.1。
主要训练验证为Qwen3.6-27B及文档集合检索,答案由模型裁判;自建500题与真实开放网络仍有距离,ViDoRe文本/页图接近且仅1.7%需视觉,不应声称全面取代OCR。
原始摘要
Complex documents such as scientific papers contain rich information in tables, figures, and equations; however, existing deep research (DR) agents typically reduce papers to OCR-extracted text. We instead ask whether vision-language models (VLMs) can enable page-native deep research: directly retrieving and reasoning over rendered document page images. We introduce DR Pixel, the first open-source end-to-end multimodal DR agent that uses rendered page images as its sole retrieval and reasoning unit, operating over a corpus of 1.06M pages. To evaluate this setting, we construct DRPixelBench, a benchmark on scientific papers requiring reasoning over text, figures, tables, and equations. Surprisingly, page-native DR already works well without training, outperforming OCR-based baselines across state-of-the-art models like GPT-5.4, Gemini-3.5-Flash, and Qwen3.6-27B, while requiring fewer tool calls. We next turn to DR training. A key challenge is that teacher trajectories for agentic supervised finetuning are noisy, mixing critical reasoning and action steps with redundant or failed attempts. To address this, we introduce counterfactual credit reweighting, a fine-grained credit assignment method that weights individual reasoning and action steps by their counterfactual contribution to the final answer. Our training improves retrieval recall by about 15 points and final accuracy by over 10 points on DRPixelBench, and likewise outperforms the untrained baseline on four existing benchmarks. On out-of-distribution ViDoRe-V3, DR Pixel comes within 1.8 points of an oracle that is given the gold pages directly. Our results suggest that the conventional OCR-then-retrieve pipeline is not only unnecessary but suboptimal: directly retrieving and reasoning over visual document pages makes complex DR both more effective and more efficient.