对比语言-图像预训练的嵌入几何
Embedding Geometries of Contrastive Language-Image Pre-Training
AI总结：
本文重新审视了CLIP的L2归一化和余弦相似度设计，提出基于欧几里得几何的变体EuCLIP，实验证明其性能匹配或超越CLIP，且能更好支持层次关系。
摘要：
自CLIP发表以来，使用InfoNCE损失进行对比预训练的方法在桥接两个或多个模态方面变得广泛流行。尽管被广泛采用，CLIP关于L2归一化和余弦相似度logit的原始设计选择却鲜少被重新审视。我们系统地实验了语言图像预训练的替代几何和softmax logits，并发现具有直观欧几里得几何的变体，即欧几里得CLIP（EuCLIP），能够匹配或超越CLIP的性能，并且在支持层次关系方面至少与更复杂的双曲替代方案一样好。
英文摘要：
Since the publication of CLIP, the approach of using InfoNCE loss for contrastive pre-training has become widely popular for bridging two or more modalities. Despite its wide adoption, CLIP's original design choices of L2 normalization and cosine similarity logit have rarely been revisited. We have systematically experimented with alternative geometries and softmax logits for language-image pre-training and identified that variants with intuitive Euclidean geometry, Euclidean CLIP (EuCLIP), match or exceed the performance of CLIP and support hierarchical relationships at least as well as more complicated hyperbolic alternative.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
