We're working on all the nasty edge cases for text-based RAG (and vision-based RAG too!)- and writing some blog posts on our vision systems soon so stay tuned 😎

VDocRAG: Retrieval-Augmented Generation over Visually-Rich Documents (CVPR 2025) TL;DR: Dataset for multi-hop questions on visual documents, self-supervised VLM pre-training, source-available license (NTT License)

When you drag select and it selects multiple columns of text at once 😔💪

Seeing is Believing? Mitigating OCR Hallucinations in Multimodal Large Language Models Zhentao He, Can Zhang, Ziheng Wu, Zhenghao Chen, Yufei Zhan, Yifan Li, Zhao Zhang, Xian Wang, Minghui Qiu

Controlled Retrieval-augmented Context Evaluation for Long-form RAG Jia-Huei Ju, Suzan Verberne, Maarten de Rijke, Andrew Yates arxiv.org/abs/2506.20051

From Web Search towards Agentic Deep Research: Incentivizing Search with Reasoning Agents arxiv.org/abs/2506.18959

Can LLMs Replace Humans During Code Chunking? The MITRE Corporation Benchmarked on legacy government code written in ALC, MUMPS, Assembly.... LLMs 20% more factual and 10% more useful than human partitioning

Taming the Untamed: Graph-Based Knowledge Retrieval and Reasoning for MLLMs to Conquer the Unknown Bowen Wang, Zhouqiang Jiang, Yasuaki Susumu, Shotaro Miwa, Tianwei Chen, Yuta Nakashima

doctors hate him

Structured Attention Matters to Multimodal LLMs in Document Understanding arxiv.org/abs/2506.21600

Ovis-U1 Technical Report arxiv.org/abs/2506.23044

Yeah document parsing is cool, but what about CAD drawing parsing... (but with documents too 🥺)

Parsing CAD drawings is one thing but parsing architectural plans takes the craziness up a notch