Videoauto R1 Adaptive Video Reasoning

Reduce inference latency and token cost in video question-answering by selectively invoking chain-of-thought reasoning. VideoAuto-R1 uses confidence-based early-exit to determine when explicit reasoning is necessary, achieving 3.3× response length reduction while maintaining state-of-the-art accuracy.

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit 4f004d5ae6

Frequently asked questions

npx get-skill add adu2021/videoauto-r1-adaptive-video-reasoning