AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Distributional Vision-Language Alignment by Cauchy-Schwarz Divergence

Wenzhe Yin, Zehao Xiao, Pan Zhou, Shujian Yu, Jiayi Shen, Jan-Jakob Sonke, Stratis Gavves

ICLR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Contrastive vision-language alignment · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add CS divergence to InfoNCE alignment (vs Eclipse L2 prior)adapter + Kandinsky decoder · MSCOCO T2IFID: 16.53 -> 12.62Improved
add CS divergence (vs Eclipse)adapter + Karlo decoder · MSCOCO T2IFID: 23.67 -> 11.27Improved
train alignment on CC3M with CS divergenceadapter · MSCOCO T2IFID: 26.73 -> 22.88Improved
train alignment on CC12M with CS divergenceadapter · MSCOCO T2IFID: 26.98 -> 22.72Improved
train alignment on LAION-HighRes 5M with CS divergenceadapter · MSCOCO T2IFID: 19.16 -> 14.79Improved
use LoRA instead of adapter (ablation)CS-Aligner Kandinsky · MSCOCO T2IFID: 12.62 -> 13.52Mixed
use LoRA instead of adapter (ablation)CS-Aligner Karlo · MSCOCO T2IFID: 11.27 -> 15.63Hurt
add 40K paired plus 80K unpaired samples for CS divergence (ablation)CS-Aligner Kandinsky · MSCOCO T2IFIDImproved
use multiple captions per image in CS divergence (ablation)CS-Aligner · MSCOCO T2IFIDImproved
add token-level alignment (ablation)CS-Aligner Kandinsky · MSCOCO T2IFIDImproved
replace CS divergence with KL divergence (ablation)CS-Aligner Kandinsky · MSCOCO T2IFID: 23.48 -> 12.62Improved
replace CS divergence with Wasserstein distance (ablation)CS-Aligner Kandinsky · MSCOCO T2IFID: 18.41 -> 12.62Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy