arXiv Machine Learning July 24, 2026 By Kisung You

Semantics at an Angle: When Cosine Similarity Works Until It Doesn't

Read the original on arXiv Machine Learning →

arXiv:2504. 16318v3 Announce Type: replace Abstract: Cosine similarity is a standard comparison rule for learned representations in information retrieval, natural language processing, computer vision, and multimodal learning.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Hugging Face Trending Papers

Jun 28

Anisotropy Decides Cosine vs. Rank Metrics for Text Embeddings

The standard way to compare two text embeddings is cosine similarity. Scattered studies report that a different metric does better, but never pin down the geometric condition that decides when, or why.

llms rag fine-tuning

arXiv Machine Learning

Jun 10

In Defense of Cosine Similarity: Normalization Eliminates the Gauge Freedom

arXiv:2602. 19393v2 Announce Type: replace Abstract: Steck, Ekanadham, and Kallus [arXiv:2403.

By Taha Bouhsine

rag

arXiv AI

Sep 15

PACE: Progressive Angular-to-Norm Contrastive Embedding

arXiv:2609.15152v1 Announce Type: cross Abstract: Multimodal embedding models encode heterogeneous inputs into a shared embedding space, enabling efficient similarity computation across modalities an...

By Yanping Li, Wei Zhou, Yawen Liu, Yibo Wang, Ke Zhu, Guangda Huzhang, Qing-Guo Chen, Zhao Xu, Jun Zhang, Wei Wei

rag fine-tuning multimodal

arXiv Computer Vision

3d ago

Spherical Interpolation for Backward-Compatible Multimodal Representations

arXiv:2609.39836v1 Announce Type: new Abstract: Contrastive vision-language models map visual and textual representations into a shared normalized embedding space, making cosine similarity the natura...

By Simone Ricci, Niccol\`o Biondi, Federico Pernici

llms rag multimodal benchmarks safety

arXiv AI

Jun 30

Multimodal Representation Alignment for Cross-modal Information Retrieval

arXiv:2506. 08774v2 Announce Type: replace-cross Abstract: Different machine learning models can represent the same underlying concept in different ways.

By Fan Xu, Luis A. Leiva

llms rag multimodal benchmarks safety

arXiv AI

Sep 1

Centering before Pruning: Lightweight Geometry Correction for Diversity-Based Visual Token Pruning in LVLMs

arXiv:2608.30263v1 Announce Type: cross Abstract: Large vision-language models (LVLMs) incur substantial inference costs due to their long and highly redundant visual-token sequences. Diversity-based...

By Shunjie Wen, Jaeyeon Lee, Dong-Wan Choi

llms efficiency multimodal benchmarks

One email a morning, machine-written

Related stories

Anisotropy Decides Cosine vs. Rank Metrics for Text Embeddings

In Defense of Cosine Similarity: Normalization Eliminates the Gauge Freedom

PACE: Progressive Angular-to-Norm Contrastive Embedding

Spherical Interpolation for Backward-Compatible Multimodal Representations

Multimodal Representation Alignment for Cross-modal Information Retrieval

Centering before Pruning: Lightweight Geometry Correction for Diversity-Based Visual Token Pruning in LVLMs