Perceptron 发布 35B 具身智能体模型 Mk1.5,请求完成速度最高提升 4.7 倍

Rohan Paul · @rohanpaul_ai · X·2026-09-26 00:13·2小时前
AI 导读

Perceptron 发布 35B 具身智能体模型 Mk1.5,可控制无人机、四足机器人、智能眼镜,并调度子智能体,无需平台特定重训练。模型针对 egocentric 视频训练,启用网页搜索、页面阅读、反向图搜和缩放工具后,MMSearch F1 从 18.2 升至 54.3;官方数据显示端到端延迟最高较 Mk1 快 4.7 倍(Chat 从 5.2 秒降至 1.1 秒)。

Rohan Paul@rohanpaul_ai
51AI 编辑部评分,满分 100

Perceptron 发布 35B 具身智能体模型 Mk1.5,请求完成速度最高提升 4.7 倍

2026-09-26 00:13· 2小时前
AI 导读

Perceptron 发布 35B 具身智能体模型 Mk1.5,可控制无人机、四足机器人、智能眼镜,并调度子智能体,无需平台特定重训练。模型针对 egocentric 视频训练,启用网页搜索、页面阅读、反向图搜和缩放工具后,MMSearch F1 从 18.2 升至 54.3;官方数据显示端到端延迟最高较 Mk1 快 4.7 倍(Chat 从 5.2 秒降至 1.1 秒)。

Perceptron released Mk1.5, a 35B embodied-agent model with up to 4.7× faster request completion.

  1. First-person video is a very different problem from ordinary video understanding.

For a robot or a pair of smart glasses, hands constantly enter the frame, manipulate objects, disappear, return, and partially occlude what matters. The camera itself is moving too.

Perceptron trained Mk1.5 specifically on egocentric video, and very interestingly the model gets a 36.1-point jump on MMSearch when tools are enabled. Moves from 18.2 to 54.3 F1 when it can use web search, page reading, reverse-image search, and zoom.

This is a useful way to think about embodied intelligence too.

A robot does not need every possible capability encoded into its weights. It needs to recognize when its current information is insufficient, pick the right external capability, use it, then continue the task.

Perceptron AIToday we're releasing Mk1.5: a new intelligence layer for embodied agents. It flies drones, controls quadrupeds, powers smart glasses, tracks objects, searches ...