EviSeek：以证据为中心的具身主动信息获取框架
EviSeek: An Evidence-Centric Agentic Framework for Embodied Active Information Seeking
一个不用训练的 agent 框架，让机器人主动找证据回答问题，在 ESI-Bench 上用 Qwen3-VL-32B 提了 7 个点还省 token。
EviSeek 是一个无需训练的框架，针对具身主动问答场景，把问题推导出的证据需求显式化为获取目标。它在 ESI-Bench 的 3,436 道题上测试，使用 Qwen3-VL-32B 时准确率比单模型 agent 提升 7.1 个百分点，使用 Gemini 3.1 Pro 时提升 3.7 个百分点。该框架同时提升了视觉支持加权准确率，并降低了平均 token 用量。
EviSeek: An Evidence-Centric Agentic Framework for Embodied Active Information Seeking
Active 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.