Story
arxiv_cs_ai ยท Nov 16, 2026 ยท paper
Source brief
ViRA: Toward Ultra-High-Resolution Visual Agents via Cross-Input Relative Policy Optimization
arxiv.orgNov 16, 2026
original source linked
In brief
Visual agents answer questions about ultra-high-resolution photos by zooming into the relevant regions. However, existing agents often answer from common sense instead of reading the evidence. Reinforcement learning (...
Feed lens
agentevaluation