ADOPD 2026: From Page Decomposition to Grounded Document Reasoning

Existing document understanding benchmarks have largely focused on locating page elements, yet real-world document intelligence requires models to reason jointly about region semantics, spatial relations, and visual structure. We present ADOPD 2026, a reasoning-oriented extension of ADOPD that turns page decomposition into spatially grounded document understanding. ADOPD 2026 enriches page primitives inherited from the ADOPD dataset with human-cleaned captions, semantic tags, and generated chain-of-thought (COT) traces grounded to document regions. Instead of treating boxes, masks, and tags as independent supervision signals, we cast text blocks, visual entities, semantic labels, bounding boxes, and polygon masks as a shared vocabulary of $\textit{visual primitives}$. This representation supports three connected capabilities. First, region-level semantic tagging asks models to identify document element types from both page context and local appearance, revealing long-tail semantic failures that standard layout benchmarks often hide. Second, unified vision-language grounding generates text regions and visual entities together with coordinates or polygonal outlines, transforming detection and segmentation outputs into structured primitives that can be reused by downstream reasoning systems. Third, current state-of-the-art models still struggle with dense counting tasks evaluated on a benchmarking dataset derived from ADOPD 2026, highlighting the need for the thinking-with-visual-primitives pipeline in document semantic understanding. By connecting page decomposition to verifiable visual-primitive reasoning, ADOPD 2026 provides a task framework that moves document understanding beyond localization toward primitives grounded document intelligence.