Ai2 @ai2.bsky.social · 18/03/2026Most VLMs point by generating text coordinates—essentially dictating numbers. It works, but it wastes tokens, breaks at high resolutions, and forces models to learn an abstract numbering system that has nothing to do with how they actually see. 110 0
Ai2 @ai2.bsky.social · 18/03/2026MolmoPoint takes a different approach. Instead of writing coordinates, the model points by selecting from the visual tokens it's already looking at—like the difference between reading out "position 347, 582" and tapping directly on a touchscreen. 120 0
Ai2 @ai2.bsky.social · 18/03/2026MolmoPoint works in three steps using special grounding tokens: 1️⃣ Pick a rough region of the image/video that contains the target 2️⃣ Zoom in to a smaller area using finer visual features 3️⃣ Pinpoint the exact pixel-level location within that area 110 0
