AI Paper Atlas

Research map / Video / Video-language understanding

Molmo2: Open Weights and Data for Vision-Language Models with Video Understanding and Grounding

Christopher Clark, Jieyu Zhang, Zixian Ma, Jae Sung Park, Rohun Tripathi, Sangho Lee, Mohammadreza Salehi, Jason Ren, Chris Dongjoo Kim, Yinuo Yang, Vincent Shao, Yue Yang and 8 more

CVPR 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 153 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
remove bidirectional attention between vision tokens (ablation)Molmo2-4B (video-only) · video benchmarksCaption F1: 32.6 -> 30.8Hurt
remove bidirectional attention between vision tokens (ablation)Molmo2-4B (video-only) · video QAQA avg.: 64.8 vs 64.4 withoutImproved
add token weighting (ablation)Molmo2-4B (video-only) · video QAQA avg.: 64.0 -> 64.8Improved
add token weighting (ablation)Molmo2-4B (video-only) · video captioningCaption F1: 34.0 -> 32.6Hurt
add Molmo2-Cap and Molmo2-QA to academic video SFT data (ablation)Molmo2-4B (video-only) · video SFTCaption F1: 4.7 -> 32.6Improved
use merged video+frame captions instead of human transcript captions only (ablation)Molmo2-4B captioner · Molmo2-CapCaption F1: 22.1 -> 33.2Improved
add video-only or frame-only captions to merged captions (ablation)Molmo2-4B captioner · Molmo2-CapCaption F1: 33.2 -> 32.3No clear effect
train jointly with images instead of video-only (ablation)Molmo2-4B · video benchmarksQA avg.: 64.8 -> 66.7Improved
predict points before counting (ablation)Molmo2-4B · Molmo2-VideoCountaccuracy: 28.1 -> 34.5Improved
upsample medium and high-count pointing examples (ablation)Molmo2-4B · Molmo2-VideoPointF1: 28.1 -> 31.8Improved
train specialized pointing-only model instead of joint model (ablation)Molmo2-4B · Molmo2-VideoPointF1: 28.4 -> 31.8Improved
add point-based single-object tracking sub-task (ablation)Molmo2-4B · Academic VOS trackingHOTA: 66.2 -> 66.7 vs 67.2 with groundingHurt

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