EviSeek: An Evidence-Centric Agentic Framework for Embodied Active Information Seeking
EviSeek is a training-free framework that makes question-derived evidence demands explicit acquisition objectives.
EviSeek is a novel framework that makes question-derived evidence demands explicit acquisition objectives, unlike prior work that leaves the informational purpose of an interaction implicit.
Before reading this…
Applications
- →Active embodied question answering
To understand this paper, make sure you know these concepts first:
- Active embodied question answeringfind papers →
- Question-derived evidence demandsfind papers →
Abstract
More Like ThisActive embodied question answering requires agents to acquire evidence through viewpoint changes and physical interaction. The central challenge is to avoid undirected interaction: taking plausible actions without specifying what evidence they should reveal for answering the question. Action plans and accumulated histories alone can leave the informational purpose of an interaction implicit. We introduce EviSeek, a training-free framework that makes question-derived evidence demands explicit acquisition objectives, linking observation-conditioned actions and selectively retained evidence to the information the agent seeks. On 3,436 ESI-Bench questions, EviSeek improves accuracy over the Single-model agent by 7.1 percentage points with Qwen3-VL-32B and 3.7 points with Gemini 3.1 Pro, while improving visual-support-weighted accuracy and reducing mean token usage. These findings support making the informational purpose of interaction explicit, with action selection and evidence retention organized around what the agent needs to learn to answer the question.