V-JEPA-LM: 1B-8B video-language models with counterfactual dynamics supervision, 4B scores 72.1

_akhaliq · x · 2026-11-17

V-JEPA-LM (1B-8B) attaches a frozen V-JEPA 2.1 encoder to an LLM and trains with counterfactual supervision so the models actually read video dynamics — direction, speed, and order — rather than relying on single-frame captions.

Related event: V-JEPA-LM: Counterfactual Supervision Teaches LLMs to Understand Video(2 posts)→

Original post →

More from Multimodal

Multimodal channel →