AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Molmo-Point: Better Pointing for VLMs with Grounding Tokens

Christopher Clark, Yue Yang, Jae Sung Park, Zixian Ma, Jieyu Zhang, Rohun Tripathi, Mohammadreza Salehi, Sangho Lee, Ranjay Krishna

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants

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 grounding tokens to tracking model (ablation)MolmoPoint-8B-Ablation · Molmo2Track overallF1: 54.7 -> 59.3Improved
add grounding tokens to tracking model (ablation)MolmoPoint-8B-Ablation · Molmo2Track overallHOTA: 55.3 -> 59.3Improved
add MolmoPoint-Track data (ablation)MolmoPoint-8B-Ablation · Molmo2Track MiscF1: 30.3 -> 37.9Improved
add rotary embeddings (ablation)Molmo2-P-Ablation · Molmo2-VCCorrect: 56.8 -> 58.0Improved
add rotary embeddings (ablation)Molmo2-P-Ablation · PixMo-PointsF1: 84.5 -> 85.2Improved
add no-more-points class (ablation)Molmo2-P-Ablation · Molmo2-VCOvercount: 10.3 -> 3.6Improved
add no-more-points class (ablation)Molmo2-P-Ablation · PointBenchAvg: 66.6 -> 67.8Improved
add sorted point generation order (ablation)Molmo2-P-Ablation · Molmo2-VCCorrect: 40.0 -> 58.0Improved
add sorted point generation order (ablation)Molmo2-P-Ablation · PointBenchAvg: 71.2 -> 67.8Hurt
use grounding tokens instead of text coordinates on natural imagesMolmoPoint-8B · PixMo-PointsF1: 85.2 -> 89.2Improved
use grounding tokens vs text coordinates for GUI fine-tuningMolmoPoint-GUI-8B · ScreenSpot-Proaccuracy: 52.3 -> 60.2Improved
add 64-crop test-time scalingMolmoPoint-GUI-8B · ScreenSpot-Proaccuracy: 60.2 -> 61.1Improved

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