DocArena: Turning Raw Documents into Controllable Training Environments for Document Search Agents
Quick Answer
DocArena introduces an automated pipeline for creating multimodal training environments for search agents, utilizing MLLM-based visual perception.
Quick Take
DocArena introduces an automated pipeline for creating multimodal training environments for search agents, utilizing MLLM-based visual perception. The system generates QA pairs from 8,336 documents across 16 domains and 49 languages, achieving superior retrieval accuracy and QA quality compared to existing methods.
Key Points
- DocArena automates document curation without human annotation, enhancing training efficiency.
- Introduces DocArena-79K with QA pairs from 8,336 documents across diverse domains.
- Decouples visual perception from policy models for improved multimodal retrieval.
- Achieves best performance on retrieval accuracy and QA quality in benchmark tests.
- Confirms effectiveness and controllability of the training environment through agent behavior analysis.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Recent methods train search agents via reinforcement learning from (question, answer, evidence) tuples without requiring expert trajectories. The tuples serve as the training environment, and whose properties directly shape what search strategies and generalization abilities the agent can develop. While prior works have made encouraging progress in improving training data quality, existing environments remain predominantly text-based and existing approaches can struggle to construct training environments that are controllable, scalable, and account for multimodal data. Given this, we propose DocArena, a fully automated data curation pipeline building on the practical need for multimodal document search and question-answering. It transforms raw document collections into training environments for search agents without any human annotation. The pipeline first structures and indexes documents through MLLM-based visual perception, then profiles and leverage the cross-page information distribution to construct reasoning-intensive QA pairs, as well as performs cascaded quality assurance operations via MLLM. We introduce DocArena-79K with QA pairs from 8,336 documents spanning 16 domains and 49 languages. We further design a Doc-Search agent infrastructure that decouples visual perception from the policy model, allowing text-based LLMs to serve as the reasoning backbone for multimodal document retrieval and QA. Under a unified evaluation framework where only the policy model differs, experiments on six multimodal document scenarios and seven text-based QA benchmarks show that agents trained on DocArena data achieve the best performance on both retrieval accuracy and QA quality. Further analysis on agent search behaviors confirms the effectiveness and controllability of the constructed training environment.
| Comments: | search agent for documents |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2606.26122 [cs.CV] |
| (or arXiv:2606.26122v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2606.26122 arXiv-issued DOI via DataCite |
Submission history
From: Samyadeep Basu [view email]
[v1]
Wed, 27 May 2026 21:21:42 UTC (1,191 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →LLM-Guided ANN Index Optimization for Human-Object Interaction Retrieval
A phase-aware LLM agent optimizes human-object interaction retrieval, outperforming Optuna TPE by 33.3% and VDTuner by 34.2% on the HICO-DET benchmark. This method enhances throughput by 15.3x over UniIR and demonstrates strong transferability across vector database management systems.