AI Paper Atlas

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

Learning multimodal dictionary decompositions with group-sparse autoencoders

Chiraag Kaushik, Davis Barch, Andrea Fanelli

ICLR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Unimodal-to-multimodal representation transfer

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 group-sparsity loss (ablation)TopK SAE (CLIP) · CIFAR-10 zero-shotaccuracy: 0.657 -> 0.808Improved
add group-sparsity loss (ablation)TopK SAE (CLIP) · CIFAR-100 zero-shotaccuracy: 0.418 -> 0.526Improved
add group-sparsity loss (ablation)TopK SAE (CLIP) · ImageNet zero-shotaccuracy: 0.303 -> 0.354Improved
add cross-modal random masking to GSAE (ablation)GSAE (CLIP) · CIFAR-10 zero-shotaccuracy: 0.808 -> 0.842Improved
add cross-modal random masking to GSAE (ablation)GSAE (CLIP) · ImageNet zero-shotaccuracy: 0.354 -> 0.373Improved
add group-sparsity loss (ablation)TopK SAE (CLAP) · GTZAN Genres zero-shotaccuracy: 0.376 -> 0.705Improved
add cross-modal random masking to GSAE (ablation)GSAE (CLAP) · GTZAN Genres zero-shotaccuracy: 0.705 -> 0.672Hurt
add cross-modal random masking to GSAE (ablation)GSAE (CLAP) · FMACaps retrievalMRR: 0.050 -> 0.061Improved
add group-sparsity loss and masking (ablation)TopK SAE (SigLIP2) · ImageNet zero-shotaccuracy: 0.451 -> 0.481Improved
add masking to GSAE (ablation)GSAE (SigLIP2) · CIFAR-10 zero-shotaccuracy: 0.922 -> 0.906No clear effect
add group-sparsity loss and masking (ablation)TopK SAE (AIMv2) · ImageNet zero-shotaccuracy: 0.443 -> 0.522Improved
add group-sparsity loss and masking (ablation)TopK SAE (MS COCO) · CIFAR-10 zero-shotaccuracy: 0.547 -> 0.820Improved

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