When Modality Gap Reduction Fails: Prediction-Level Hubness in CLIP

2026-09-01

AI Summary

本文探讨了减少CLIP中图像-文本模态差距并不总能提高零样本分类准确率的问题，通过分析决策结构提出预测级集中现象，并建议评估方法应考虑对下游预测结构的影响。

Abstract

Reducing the modality gap between image and text representations in CLIP is widely expected to improve cross-modal alignment and downstream performance. However, a smaller average image-text gap does not necessarily lead to consistent accuracy gains. We analyze this mismatch from the perspective of the decision structure in zero-shot classification, i.e. selecting the most similar class-text prototype for an input image. Zero-shot accuracy depends not only on average image--text alignment, but also on class-wise decision margins. Using Linear correction as an analytically tractable case, we show that modality gap correction can alter the relative decision structure among classes and cause predictions to concentrate on a small subset of classes. We refer to this output-space failure mode as prediction-level hubness. Furthermore, experiments across multiple datasets show that accuracy degradation under gap correction is consistently associated with increased prediction concentration, both for Linear correction and for learning-based correction methods. This provides a systematic explanation of why modality gap reduction does not consistently improve CLIP zero-shot accuracy from the perspective of downstream decision structure. Our results suggest that gap correction should be evaluated not only by average alignment, but also by its impact on downstream prediction structure.

Problem

• modality gap
• zero-shot classification
• prediction-level hubness
• cross-modal alignment

Innovation

• modality gap
• zero-shot classification
• prediction-level hubness
• decision structure