AI Paper Atlas

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

No Hard Negatives Required: Concept Centric Learning Leads to Compositionality without Degrading Zero-shot Capabilities of Contrastive Models

Hai X. Pham, David T. Hoffmann, Ricardo Guerrero, Brais Martinez

CVPR 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Self-distillation for vision-language pretraining · 0 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
fine-tune on DreamLIP CC3M with standard loss only (ablation)SigLIP ViT-B/16 · SugarCrepe averageaccuracy: 79.5 -> 81.7Improved
add noun-phrase concept loss (npc) (ablation)C2LIP · SugarCrepe averageaccuracy: 81.7 -> 85.3Improved
add cross-attended concept-aware loss (xac) (ablation)C2LIP · SugarCrepe averageaccuracy: 85.3 -> 85.6Improved
add npc loss (ablation)C2LIP · Flickr8K retrievalRecall@5: 95.2 -> 96.3Improved
add xac loss (ablation)C2LIP · SugarCrepe Replace attributeaccuracy: 89.7 -> 89.2Hurt
use C2LIP objectives instead of plain CC3M fine-tuningC2LIP · ImageNet1K zero-shotaccuracy: 75.9 -> 73.5Hurt
use C2LIP objectives instead of plain CC3M fine-tuningC2LIP · SugarCrepe++ Swap TOTaccuracy: 36.6 -> 44.2Improved
use C2LIP objectives instead of plain CC3M fine-tuningC2LIP · Overall averageaverage score: 71.5 -> 75.0Improved
use C2LIP vision encoder in LLaVALLaVA · SugarCrepe averageaccuracy: 83.5 -> 83.9Improved
use C2LIP vision encoder in LLaVALLaVA · SugarCrepe++ averageaccuracy: 68.1 -> 69.7Improved
use lambda_npc 1, lambda_xac 0.01 instead of 0.5 and 0.5 (ablation)C2LIP · SugarCrepe averageaccuracy: 84.6 -> 85.6Improved

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