Single-cell RNA sequencing profiles cellular heterogeneity at atlas scale, making automated annotation essential. However, target datasets often contain novel cell types missing from incomplete references. We present scOLAR, an ontology-guided open-set framework that learns prototypes over the Cell Ontology and uses both reference and target expression to annotate known classes while detecting unfamiliar populations. Guided by ontology hierar-chies and decision-boundary regularization, scOLAR penalizes coarse-lineage misclassification and groups novel cells without requiring predefined cluster counts. Across benchmarks, scOLAR achieves a novelty-detection AUROC of 0.9726 and an average precision of 0.9871, enabling structured post-hoc lineage-level interpretation of populations absent from the reference.
Pathology vision-language models (VLMs) have recently progressed rapidly and are commonly evaluated by answer accuracy on pathology VQA benchmarks. However, we dig into current evaluations and identify three overlooked issues: 1) Visual evidence is not always necessary. For instance, Gemini-3-Pro achieves 53.5% average accuracy across 5 VQA benchmarks without any visual input. 2) Domain training can improve accuracy without proportional gains in visual binding. Compared with Qwen2.5-VL-7B, Patho-R1-7B exhibits a 5.8-point lower multimodal gain and a 3.7-point lower attention IoU. 3) Entity-level attention is diffuse and weakly query-specific. On PathVG, attention maps remain highly correlated across different entity queries. These issues can lead to substantial misjudgments of pathology VLMs'actual multimodal capabilities. To this end, we present PathBind, a benchmark comprising 2,600 samples: PathBind-VQA with 1,500 questions across six dimensions, PathBind-PTA with 600 questions from a private pathology teaching atlas, and PathBind-Grounding with 500 expert-curated region-level samples. Each component undergoes task-specific automated filtering and expert review to reduce textual shortcuts and improve entity-region correspondence. We evaluate 18 representative VLMs on VQA samples of PathBind and five existing pathology VQA benchmarks, and further evaluate 10 VLMs on PathBind-Grounding and PathVG. Results show that current pathology VLMs still exhibit a substantial gap between answer-side performance and visual-semantic binding.
Chengyang Zhang, Wenchuan Zhang, Bo Li et al.· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
