1 paper · 1 filter
Yinan Zhou, Haokun Lin, Yichen Wu +7
Large multimodal models have achieved strong reasoning on complex visual tasks, but their inference efficiency is often restricted by long chains of thought. A promising solution i…