paper-with-me

홈 › Papers

Cluster-Aware Neural Collapse Prompt Tuning for Long-Tailed Generalization of Vision-Language Models

2026-05-12 · Boyang Guo, Liang Li, Lin Peng, Yuhan Gao, Xichun Sheng, Chenggang Yan arxiv

Prompt learning has emerged as an efficient alternative to fine-tuning pre-trained vision-language models (VLMs). Despite its promise, current methods still struggle to maintain tail-class discriminability when adapting to class-imbalanced datasets. In this work, we propose cluster-aware neural collapse prompt tuning (CPT), which enhances the discriminability of tail classes in prompt-tuned VLMs without sacrificing their overall generalization. First, we design a cluster-invariant space by mining semantic assignments from the pre-trained VLM and mapping them to prompt-tuned features. This computes cluster-level boundaries and restricts the constraints to local neighborhoods, which reduces interference with the global semantic structure of the pre-trained VLM. Second, we introduce neural-collapse-driven discriminability optimization with three losses: textual Equiangular Tight Frame (ETF) separation loss, class-wise convergence loss, and rotation stabilization loss. These losses work together to shape intra-cluster geometry for better inter-class separation and intra-class alignment. Extensive experiments on 11 diverse datasets demonstrate that CPT outperforms SOTA methods, with stronger performance on long-tail classes and good generalization to unseen classes.

📄 PDF Abstract BibTeX arXiv:2605.11939

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Cluster-Aware Prompt Ensemble Learning for Few-Shot Vision-Language Model Adaptation

2025-10-10 · Zhi Chen, Xin Yu, Xiaohui Tao, Yan Li 외 arxiv

Vision-language models (VLMs) such as CLIP achieve zero-shot transfer across various tasks by pre-training on numerous image-text pairs. These models often benefit from using an ensemble of context prompts to represent a…

Ensemble Learning

CRAFT: Cost-aware Refinement And Front-aware Tuning of Prompts

2026-06-03 · Shanu Kumar, Shubhanshu Khandelwal, Akhila Yesantarao Venkata, Parag Agrawal 외 arxiv

Prompts tuned for accuracy often grow long, raising inference cost on every model call. The best accuracy-cost trade-off depends on the task and the budget, so prompt optimization is a search over the Pareto front of acc…

Exploring Embedding Priors in Prompt-Tuning for Improved Interpretability and Control

2024-12-24 · Sergey Sedov, Sumanth Bharadwaj Hachalli Karanam, Venu Gopal Kadamba

Prompt-Tuning is an efficient method for adapting pre-trained language models to new tasks with minimal computational overhead by modifying prompt embeddings. In this work, we investigate how crucial the phenomenon of em…

Question Answering

CAPT: Class-Aware Prompt Tuning for Federated Long-Tailed Learning with Vision-Language Model

2025-03-10 · Shihao Hou, Xinyi Shang, Shreyank N Gowda, Yang Lu 외

Effectively handling the co-occurrence of non-IID data and long-tailed distributions remains a critical challenge in federated learning. While fine-tuning vision-language models (VLMs) like CLIP has shown to be promising…

Federated LearningLanguage ModelingLanguage ModellingPrompt Learning

Understanding Prompt Tuning for V-L Models Through the Lens of Neural Collapse

2023-06-28 · Didi Zhu, Zexi Li, Min Zhang, Junkun Yuan 외

Large-scale vision-language (V-L) models have demonstrated remarkable generalization capabilities for downstream tasks through prompt tuning. However, the mechanisms behind the learned text representations are unknown, l…