TY  - RPRT
TI  - Fewer Tokens, More Answers: Verifiable-Reward Tuning of an Ornith-Style VLM for Multi-View Driving QA
AU  - T N Rickesh
AU  - Pranav Vinodh
AU  - Sowmya Kamath S
PY  - 2026
UR  - https://arxiv.org/abs/2611.00833
ID  - 2611.00833
ER  -