AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

VL-JEPA: Joint Embedding Predictive Architecture for Vision-language

Delong Chen, Mustafa Shukor, Théo Moutakanni, Willy Chung, Lei Yu, Tejaswi Kasarla, Allen Bolourchi, Yann LeCun, Pascale Fung

ICLR 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Language-grounded scene text and structure · 68 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
predict target embeddings instead of tokens (strictly matched training)VL-JEPA vs VLM, 15M samples · zero-shot video captioningCIDEr: 7.1 -> 14.8Improved
predict target embeddings instead of tokens (strictly matched training)VL-JEPA vs VLM, 15M samples · zero-shot video classificationtop-5 accuracy: 27.2 -> 41.0Improved
add caption pretraining stage before SFT (ablation)VL-JEPA · ablation, classification and retrievalaccuracy / recall@1: -21.7 classification, -17.3 retrieval without itImproved
embedding-guided selective decoding instead of uniform decoding (ablation)VL-JEPA · EgoExo4D streamingdecoding operations at matched CIDEr: ~2.85x fewer decodes (0.35 Hz matches 1 Hz)Improved
average pooling of embeddings before decoding (ablation)VL-JEPA selective decoding · EgoExo4D streamingCIDErImproved
Y-Encoder learning rate multiplier 1.00 instead of 0.05 (ablation)VL-JEPA · ablation, classificationaccuracy: 27.3 -> 23.7Hurt
Y-Encoder learning rate multiplier 0.00 (frozen) instead of 0.05 (ablation)VL-JEPA · ablation, classificationaccuracy: 27.3 -> 20.0Hurt
cosine, L1 or L2 loss instead of InfoNCE (ablation)VL-JEPA · ablation, classification and retrievalaccuracy / recall@1: classification 23.3 -> 16.5 (cosine), 13.5 (L2)Hurt
cosine loss instead of InfoNCE (ablation)VL-JEPA · ablation, VQAaccuracy: 44.3 -> 46.6Improved
more predictor layers (Layer 0-16 vs 8-16) (ablation)VL-JEPA · ablation, VQAaccuracy: 42.5 -> 45.5Improved
causal attention instead of bi-directional attention in predictor (ablation)VL-JEPA · ablation, VQAaccuracy: 42.5 -> 40.6Hurt
random initialization instead of Llama-3 initialization of predictor (ablation)VL-JEPA · ablation, VQA vs classificationaccuracy: VQA 42.5 -> 40.6, classification 27.3 -> 28.1Mixed

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