Harnessing Image Question Dependence for Better VLM Test-time Reinforcement Learning
The proposed TTIQ, a test-time reinforcement learning framework that harnesses image-question dependence for better vlm adaptation, combines image and question dependence with calibrated confidence to construct a response-level reward that favors jointly grounded responses.
Xinrui He, Ting-Wei Li, Jun-Ting Wang et al.
· 0 citations