Skip to content
Hot eventLive

何恺明团队提出多模态视觉原生Harness框架VISTA

1 reports1 sources5 hr ago updated

Get the story

AI overview

2026年10月11日，量子位报道，何恺明团队在论文《VISTA: A Visual Harness for Reasoning in an Interactive World》中提出视觉原生Harness框架VISTA。该框架面向交互式世界中的推理任务，让现有多模态模型无需额外训练即可保存原始画面，并在推理过程中回看、放大和检查细节。目前公开信息仅涉及该论文与框架的基本能力，尚无后续验证或应用进展报道。

Generated from reports · updated 3 hr ago

LatestOct 11
何恺明团队提出VISTA框架，支持现有多模态模型免训练保存原始画面并回看、放大、检查细节。

Timeline

Follow the coverage from different angles.

Oct 11, 2026
  1. 量子位精选
    何恺明团队提出多模态视觉原生Harness框架VISTA

    何恺明团队在论文《VISTA: A Visual Harness for Reasoning in an Interactive World》中提出视觉原生Harness框架VISTA，让现有多模态模型无需额外训练即可保存原始画面，并在推理时回看、放大和检查细节。

Heat trend

Current heat 9·Comparable peak 9（Oct 11）·Comparable change over 24 hours –

02.557.510Oct11Oct11Oct11Oct11

The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.