PathAgentBench: Benchmarking Evidence-Seeking Vision-Language Models on Whole-Slide Pathology Image

Whole-slide image (WSI) diagnosis requires identifying diagnostically relevant regions, examining them across magnifications, and integrating multi-scale evidence. However, most existing pathology benchmarks evaluate models on pre-cropped patches or pre-extracted slide features, leaving their ability to acquire evidence directly from gigapixel WSIs largely untested. We introduce PathAgentBench, a benchmark for evaluating evidence-seeking vision-language models (VLMs) across four complementary capabilities: image-to-text matching for evidence interpretation, text-to-image retrieval for evidence verification, diagnostic-region localization for evidence acquisition, and multi-scale reasoning for evidence integration. The benchmark is organized as a diagnostic tree that links nested regions across magnifications with scale-specific findings and path-level diagnoses. It contains 1,822 TCGA WSIs and 17,135 diagnostic paths annotated by ten board-certified pathologists. An additional private cohort of 190 breast cancer WSIs with detailed annotations is used to evaluate autonomous whole-slide exploration. We evaluate 20 general-purpose, medical, and pathology-specialized models. Leading open-weight models achieve over 93% accuracy in multi-scale reasoning and over 50% accuracy in both cross-modal matching tasks. In contrast, diagnostic-region localization remains challenging: the best text-guided mean intersection-over-union is below 0.09, underperforming a simple center-based heuristic. During autonomous exploration, the unconditional hit rate decreases from 0.522 at low magnification to 0.185 at intermediate magnification and 0.020 at high magnification. These results reveal a pronounced gap between reasoning over curated evidence and acquiring that evidence directly from WSIs. PathAgentBench provides a unified framework for measuring and improving evidence-seeking pathology models.

PathRWKV: Enabling Whole Slide Prediction with Recurrent-Transformer

Pathological diagnosis plays a critical role in clinical practice, where the whole slide images (WSIs) are widely applied. Through a two-stage paradigm, recent deep learning approaches enhance the WSI analysis with tile-level feature extracting and slide-level feature modeling. Current Transformer models achieved improvement in the efficiency and accuracy to previous multiple instance learning based approaches. However, three core limitations persist, as they do not: (1) robustly address the modeling on variable scales for different slides, (2) effectively balance model complexity and data availability, and (3) balance training efficiency and inference performance. To explicitly address them, we propose a novel model for slide modeling, PathRWKV. Via a recurrent structure, we enable the model for dynamic perceptible tiles in slide-level modeling, which novelly enables the prediction on all tiles in the inference stage. Moreover, we employ linear attention instead of conventional matrix multiplication attention to reduce model complexity and overfitting problem. Lastly, we hinge multi-task learning to enable modeling on versatile tasks simultaneously, improving training efficiency, and asynchronous structure design to draw an effective conclusion on all tiles during inference, enhancing inference performance. Experimental results suggest that PathRWKV outperforms the current state-of-the-art methods in various downstream tasks on multiple datasets. The code and datasets are publicly available.

UnPuzzle: A Unified Framework for Pathology Image Analysis

Pathology image analysis plays a pivotal role in medical diagnosis, with deep learning techniques significantly advancing diagnostic accuracy and research. While numerous studies have been conducted to address specific pathological tasks, the lack of standardization in pre-processing methods and model/database architectures complicates fair comparisons across different approaches. This highlights the need for a unified pipeline and comprehensive benchmarks to enable consistent evaluation and accelerate research progress. In this paper, we present UnPuzzle, a novel and unified framework for pathological AI research that covers a broad range of pathology tasks with benchmark results. From high-level to low-level, upstream to downstream tasks, UnPuzzle offers a modular pipeline that encompasses data pre-processing, model composition,taskconfiguration,andexperimentconduction.Specifically, it facilitates efficient benchmarking for both Whole Slide Images (WSIs) and Region of Interest (ROI) tasks. Moreover, the framework supports variouslearningparadigms,includingself-supervisedlearning,multi-task learning,andmulti-modallearning,enablingcomprehensivedevelopment of pathology AI models. Through extensive benchmarking across multiple datasets, we demonstrate the effectiveness of UnPuzzle in streamlining pathology AI research and promoting reproducibility. We envision UnPuzzle as a cornerstone for future advancements in pathology AI, providing a more accessible, transparent, and standardized approach to model evaluation. The UnPuzzle repository is publicly available at https://github.com/Puzzle-AI/UnPuzzle.