Treevgr Visual Grounded Reasoning Evidence

Train vision-language models to produce visually grounded reasoning by enforcing traceable evidence via bounding box localization, using a novel benchmark (TreeBench) and RL-based training with dual IoU rewards for both recall and precision.

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit 20d0301c53

Frequently asked questions

npx get-skill add adu2021/treevgr-visual-grounded-reasoning-evidence