AI Paper Atlas

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

Text-Conditional JEPA for Learning Semantically Rich Visual Representations

Chen Huang, Xianhang Li, Vimal Thilak, Etai Littwin, Joshua M. Susskind

ICML 2026 · Vision-language and multimodal › Human and embodied modeling · approach: 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
add text conditioning to I-JEPA (ViT-B/16)I-JEPA · ImageNet-1k linear probingtop-1 accuracy: 72.9 -> 75.8Improved
add text conditioning to I-JEPA (ViT-L/16)I-JEPA · ImageNet-1k linear probingtop-1 accuracy: 77.5 -> 79.6Improved
add text conditioning to I-JEPA (ViT-H/14)I-JEPA · ImageNet-1k linear probingtop-1 accuracy: 79.3 -> 80.4Improved
add text conditioning to I-JEPA (ViT-H/14)I-JEPA · iNat18 linear probingaccuracy: 47.6 -> 54.8Improved
add text conditioning to I-JEPA (ViT-H/14)I-JEPA · ADE20k linear segmentationmIoU: 36.9 -> 39.5Improved
use MaxPool instead of AvgPool to fuse multi-caption features (ablation)TC-JEPA ViT-L/16 · ImageNet-1k linear classificationtop-1 accuracy: 79.3 -> 79.6Improved
use MaxPool instead of AvgPool to fuse multi-caption features (ablation)TC-JEPA ViT-L/16 · ADE20k linear segmentationmIoU: 38.4 -> 38.8Improved
use AttentionPool instead of MaxPool for caption fusion (ablation)TC-JEPA ViT-L/16 · ADE20k linear segmentationmIoU: 38.8 -> 39.0No clear effect
add sparse and consistency regularization losses on patch-word similarities (ablation)TC-JEPA · IN-1k pretraining ablationdownstream performanceImproved
increase number of synthetic captions N up to 8 (ablation)TC-JEPA ViT-L/16 · IN-1k and YFCC15M pretrainingdownstream performanceImproved
use diverse synthetic captions instead of the single raw caption (N=1) (ablation)TC-JEPA ViT-L/16 · YFCC15M pretrainingdownstream performanceImproved
scale loss coefficients lambda and beta by r in [0.5, 2.5] (ablation)TC-JEPA ViT-L/16 · IN-1k and YFCC15M pretrainingdownstream performanceNo clear effect

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