📄 中文摘要
大规模多模态对比学习在学习丰富且可迁移的表征方面取得了显著成功，但仍然受到特征维度均匀处理和忽视学习特征内在光谱结构的根本限制。实证证据表明，高维嵌入往往会崩溃为狭窄的锥形，任务相关的语义集中在一个小子空间内，而大多数维度则被噪声和虚假相关性占据。这种光谱不平衡和纠缠削弱了模型的泛化能力。提出了一种新的框架——光谱解耦与增强（SDE），旨在弥合嵌入空间几何与特征表示之间的差距。
光谱解耦与增强：一种双域对比框架用于表征学习
📄 English Summary
Spectral Disentanglement and Enhancement: A Dual-domain Contrastive Framework for Representation Learning
Large-scale multimodal contrastive learning has achieved remarkable success in learning rich and transferable representations. However, it is fundamentally limited by the uniform treatment of feature dimensions and the neglect of the intrinsic spectral structure of the learned features. Empirical evidence shows that high-dimensional embeddings often collapse into narrow cones, concentrating task-relevant semantics in a small subspace while the majority of dimensions are occupied by noise and spurious correlations. This spectral imbalance and entanglement undermine model generalization. The proposed Spectral Disentanglement and Enhancement (SDE) framework aims to bridge the gap between the geometry of the embedded space and feature representation.