[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82407-en":3,"doc-seo-82407-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82407,13056703020460,"Valentina","https://ap-avatar.wpscdn.com/avatar/be000253dac470eee5d?_k=1778207105932848923",8,"Research & Report","TCLA Training-Free Class-wise Logit Adaptation for Medical Vision-Language Models","Medical Vision–Language Models (VLMs) achieve strong zero-shot results, but performance drops on out-of-distribution (OOD) medical data because of domain shifts and class bias from large-scale pretraining. The paper introduces TCLA, a training-free few-shot adaptation method for medical VLMs that remains fast and model-agnostic. By adapting inference logits from a small support set, TCLA improves inter-class deconfusion and reduces domain shift. Experiments on nine datasets across X-ray, ultrasound, MRI, CT, and histopathology show consistent OOD gains.","arXiv :2607 .09562v1 [ cs .CV] 10 Jul 2026  \nTCLA: Training-Free Class-wise Logit Adaptation for Medical Vision-Language Models  \nTianyou Jianga , Ziyu Zhoub  \na University of Bern, Bern, Switzerland b Shanghai Jiao Tong University, Shanghai, China  \nAbstract  \nMedical Vision–Language Models (VLMs) exhibit strong zero-shot performance, yet their effectiveness still declines on out-of-distribution (OOD) data due to domain shifts and class bias inherited from large-scale pretraining. Existing few-shot adaptation methods typically introduce additional trainable components, which can be unstable in extremely low-data regimes (e.g., 1-shot), and lack robustness on different medical data. We present TCLA, a purely training-free few-shot adaptation method for Medical VLMs, which is fast and model-agnostic. TCLA corrects inference logits based on a small set of support samples, boosting pretrained VLMs performance by improving inter-class deconfusion and reducing domain shift. Extensive experiments on nine datasets across multiple medical imaging modalities including X-ray, Ultrasound, MRI, CT, Histopathology, demonstrate that TCLA consistently improves OOD performance of Medical VLMs and, in most of cases, outperforms existing training-based adaptation methods.  \nKeywords:  \nMedical VLMs, Few-shot Adaptation, Training-free  \nEmail address: [tianyou.jiang.research@gmail.com](tianyou.jiang.research@gmail.com) (Tianyou Jiang)  \n1. Introduction  \nVision and Language Models (VLMs) have attracted increasing attention following the success of CLIP (Contrastive Language Image Pretraining), which maps images and text into a shared embedding space [1] . The growing availability of large-scale image–caption datasets in the medical domain has further enabled Vision–Language Pretraining (VLP) for medical applications. Several Medical CLIP variants, including PubMedCLIP [7], MedCLIP [8], BioMedCLIP [9], and BIOMEDICA [10], have demonstrated strong zeroshot (ZS) performance on downstream medical imaging tasks, such as disease diagnosis [2], medical image segmentation [3], and radiology report generation [4] .  \nNevertheless, real-world medical data are often out-of-distribution (OOD), i.e., unseen disease categories present different image distributions due to variations in hospitals, scanners, or patient populations. Medical CLIPbased models can exhibit poor Zero-Shot performance when processing OOD datasets due to domain shifts and inherent class imbalance in their large-scale pretraining dataset [5][29] . A recent comparative study between generalpurpose CLIP and Medical CLIP reported that Medical CLIP models may offer only limited advantages under transfer learning settings on OOD benchmarks [5] .  \nVarious few-shot learning (FS) methods have been proposed to adapt general VLMs for improved performance. These approaches typically freeze the image and text encoders, introduce lightweight adaptive modules, and employ strategies such as Prompt Learning (PL) [11, 13], Feature Adaptation  \n(FA) [14, 15], or Logit Adaptation (LA) [16, 17, 36] . However, rapid adaptation with extremely limited labeled data has been reported to be unstable under out-of-distribution (OOD) conditions. A recent comparative analyze of few-shot adaptation methods for medical VLMs shows that conventional Linear Probing (LP) [20] can outperform more complex adaptation strategies in certain low-shot regimes [19] . Suggesting that adaptive parameters become difficult to reliably optimize and adapt Medical VLMs when only a handful of training samples (e.g., 1-shot) are available. In such scenarios, training-free approaches may offer advantages in stability, computational efficiency, and deployment simplicity.  \nMotivated by these challenges, we propose TCLA (Training-free Classwise Logit Adaptation), a training-free method that adapts the logits of medical VLMs for downstream diagnostic tasks. TCLA computes class-wise, layer-adaptive prototypes from few-shot support sets, constr","cbCaigo7a6rkn58s","https://ap.wps.com/l/cbCaigo7a6rkn58s","pdf",1294492,1,26,"English","en",105,"# Introduction\n# Related Work\n## Vision-Language Pre-trained Models","[{\"question\":\"What problem does TCLA address for medical vision-language models?\",\"answer\":\"TCLA addresses the performance decline of medical VLMs on out-of-distribution (OOD) data caused by domain shifts and class bias inherited from large-scale pretraining.\"},{\"question\":\"How does TCLA adapt medical VLMs without training?\",\"answer\":\"TCLA computes class-wise, layer-adaptive prototypes from a small support set, builds prototype-based correction bases, and uses a closed-form residual mapping to adjust zero-shot logits without updating model parameters.\"},{\"question\":\"What evidence supports TCLA’s effectiveness?\",\"answer\":\"Experiments on nine datasets across multiple modalities (X-ray, ultrasound, MRI, CT, histopathology) show consistent improvements in OOD performance, and in most cases outperform existing training-based adaptation methods.\"}]",1784180164,66,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"tcla-training-free-class-wise-logit-adaptation-for-medical-vision-language-models","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/tcla-training-free-class-wise-logit-adaptation-for-medical-vision-language-models/82407/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does TCLA address for medical vision-language models?","Question",{"text":74,"@type":75},"TCLA addresses the performance decline of medical VLMs on out-of-distribution (OOD) data caused by domain shifts and class bias inherited from large-scale pretraining.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does TCLA adapt medical VLMs without training?",{"text":79,"@type":75},"TCLA computes class-wise, layer-adaptive prototypes from a small support set, builds prototype-based correction bases, and uses a closed-form residual mapping to adjust zero-shot logits without updating model parameters.",{"name":81,"@type":72,"acceptedAnswer":82},"What evidence supports TCLA’s effectiveness?",{"text":83,"@type":75},"Experiments on nine datasets across multiple modalities (X-ray, ultrasound, MRI, CT, histopathology) show consistent improvements in OOD performance, and in most cases outperform existing training-based adaptation methods.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]