SoTA with Less: MCTS-Guided Sample Selection for Data-Efficient Visual Reasoning Self-Improvement

NeurIPSSpotlight2025

Authors
Xiyao Wang, Zhengyuan Yang, Chao Feng, Hongjin Lu, Linjie Li, Chung-Ching Lin, Kevin Lin, Furong Huang, Lijuan Wang
Venue
NeurIPS 2025
Track
Spotlight

TL;DR

We introduce ThinkLite-VL, a family of visual reasoning models that achieve state-of-the-art (SoTA) performance using an order of magnitude fewer training samples, relying purely on reinforcement fine…

Opening excerpt from the authors’ abstract. source

Read the paper

Topics

efficient reasoning

Related NeurIPS papers

← All NeurIPS 2025 Spotlight papers · Browse the whole archive