AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Efficient Discriminative Joint Encoders for Large Scale Vision-Language Reranking

Mitchell Keren Taraday, Shahaf Wagner, Chaim Baskin

ICLR 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Recaptioning and text-supervised pretraining · 1 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 EDJE re-ranker on top of embedding modelCLIP ViT-B/16 · Flickr30K zero-shot text-to-imageR@1: 62.1 -> 76.8Improved
add EDJE re-ranker on top of embedding modelSigLIP2 ViT-L/16 384 · Flickr30K zero-shot text-to-imageR@1: 82.3 -> 87.8Improved
add EDJE re-ranker on top of embedding modelDFN ViT-L/14 · Flickr30K zero-shot text-to-imageR@1: 75.1 -> 77.5Improved
add masked language modeling to ITM (ablation)EDJE local, SigLIP2 L 384 · Flickr30KR@1: 82.3 -> 85.5Improved
add text-embedding recovery (ITC) to ITM+MLM (ablation)EDJE local, SigLIP2 L 384 · Flickr30KR@1: 85.5 -> 87.8Improved
add local-to-compressed distillation (ablation)EDJE Compressed-64 · Flickr30KR@1: 83.8 -> 86.9Improved
use learned attention adapter instead of striding (ablation)EDJE 64 tokens · Flickr30K zero-shot text-to-imageR@1: 83.24 -> 86.9Improved
use learned attention adapter instead of k-means clustering (ablation)EDJE 64 tokens · Flickr30K zero-shot text-to-imageR@1: 85.66 -> 86.9Improved
use learned attention adapter instead of attention pruning (ablation)EDJE 64 tokens · Flickr30K zero-shot text-to-imageR@1: 82.4 -> 86.9Improved
mine negatives with CLIP L instead of SigLIP2 (ablation)SigLIP2-based EDJE · Flickr30K zero-shot text-to-imageR@1: 87.8 -> 81.5Hurt
mine negatives with SigLIP2 L instead of CLIP (ablation)CLIP-based EDJE · Flickr30K zero-shot text-to-imageR@1: 81.9 -> 81.8No clear effect
compress 576 vision tokens to 64 with adapterEDJE SigLIP2 L · Flickr30K zero-shot text-to-imageR@1: 87.8 -> 86.9Hurt

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