VISTA: Internalizing Collective Visual Experience via On-Policy Distillation for Active Multimodal Agents

Paper recorded by Signals 4 on 2026-09-29 in cs.CV. Abstract reproduced from arXiv; link to the original below.

Published 2026-09-29 on arXiv · recorded by Signals 4 on 2026-09-30

Category: cs.CV · 计算机视觉 · first seen 2026-09-30

Abstract

Active multimodal agents use visual tools to acquire task-relevant evidence while reasoning. Although reinforcement learning samples multiple interaction trajectories per input, outcome-based objectives primarily use the group to estimate scalar advantages, leaving complementary visual discoveries underused. We introduce VISTA, which internalizes collective visual experience through on-policy dist

Read on arXiv →