{"slug": "tap-path-task-adaptive-structural-and-token-pruning-for-efficient-and-pathology", "title": "TAP-Path: Task-Adaptive Structural and Token Pruning for Efficient and Trustworthy Pathology Foundation Models", "summary": "Researchers proposed TAP-Path, a task-adaptive compression framework that prunes a pretrained Virchow2 encoder, retaining 24 of 32 transformer blocks and 70% of patch tokens, reducing parameters by 24.96% (from 631.24M to 473.70M) and compute by 35.20% (from 340.13G to 220.40G FLOPs). On a 32-class histopathology benchmark, TAP-Path achieved 87.98% test accuracy, surpassing full Virchow2 (86.89%) and UNI2-h (87.67%), with a Brier score of 0.1800 and failure-detection AUROC of 0.9047. External evaluation on 433 CPTAC samples yielded 91.22% accuracy, demonstrating improved accuracy-efficiency trade-offs while preserving reliability.", "body_md": "arXiv:2609.04071v1 Announce Type: cross\nAbstract: Pathology foundation models improve transferable representation learning for histopathology, but recent gains often rely on encoders with hundreds of millions of parameters and high inference cost. We propose TAP-Path, a task-adaptive compression framework that directly restructures a pretrained Virchow2 encoder rather than distilling it into a separate student. TAP-Path combines validation-driven transformer-block selection, physical removal of redundant blocks, input-adaptive patch-token pruning, multi-depth feature recovery, and a lightweight gated task head. The final model retains 24 of 32 transformer blocks and 70% of patch tokens after pruning, reducing encoder parameters by 24.96% (631.24M to 473.70M) and analytical encoder compute by 35.20% (340.13G to 220.40G FLOPs). Across three task-head optimization seeds, TAP-Path achieved $87.98 \\pm 0.067%$ test accuracy, $81.26 \\pm 0.49%$ balanced accuracy, and $82.38 \\pm 0.48%$ macro-F1 on a 32-class histopathology benchmark, compared with 86.89% for full Virchow2 and 87.67% for UNI2-h. TAP-Path achieved a Brier score of $0.1800 \\pm 0.0005$ and failure-detection AUROC of $0.9047 \\pm 0.0060$. A validation-only rare-aware objective improved rare-class balanced accuracy in a secondary operating analysis. Frozen external evaluation on 433 CPTAC samples yielded $91.22 \\pm 0.83%$ accuracy and $91.10 \\pm 0.81%$ balanced accuracy. These results show that task-adaptive structural and token sparsification can improve the accuracy-efficiency trade-off of large pathology foundation models while preserving reliability under internal and external evaluation.", "url": "https://wpnews.pro/news/tap-path-task-adaptive-structural-and-token-pruning-for-efficient-and-pathology", "canonical_source": "https://www.machinebrief.com/news/tap-path-task-adaptive-structural-and-token-pruning-for-effi-j34z", "published_at": "2026-09-04 04:00:00+00:00", "updated_at": "2026-09-04 04:52:27.473936+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-research"], "entities": ["TAP-Path", "Virchow2", "UNI2-h", "CPTAC"], "alternates": {"html": "https://wpnews.pro/news/tap-path-task-adaptive-structural-and-token-pruning-for-efficient-and-pathology", "markdown": "https://wpnews.pro/news/tap-path-task-adaptive-structural-and-token-pruning-for-efficient-and-pathology.md", "text": "https://wpnews.pro/news/tap-path-task-adaptive-structural-and-token-pruning-for-efficient-and-pathology.txt", "jsonld": "https://wpnews.pro/news/tap-path-task-adaptive-structural-and-token-pruning-for-efficient-and-pathology.jsonld"}}