Flex-π: Multi-Stream World-Action Model

Loading video

Jointly predicts RGB, 3D pointmaps, DINO semantics with actions; one checkpoint deploys as VLA, full WAM, or anything in between.

VLA 世界模型 WAM 双臂 Flex-π Twitter

Category: research

Author: @GeYan_21

Date: 2026-08-12T00:00:00

Duration: 50.466s

Source: https://x.com/GeYan_21/status/2087572031567265972

Reference: https://github.com/geyan21/flex-pi

Flex-π: compute-flexible robot foundation model

Flex-π: compute-flexible robot foundation model

@yokoshaan · 5 views · 2026-08-14

机器人基础模型VLA世界模型

Galaxea G0.5: One Autoregressive VLA for Reasoning and Action

Galaxea G0.5: One Autoregressive VLA for Reasoning and Action

@robotsdigest · 13 views · 2026-08-13

Dyna-2 World-Action Model by Dyna Robotics

Dyna-2 World-Action Model by Dyna Robotics

@TheHumanoidHub · 5 views · 2026-08-10

世界模型世界-动作模型WAM

τ0-VLA: Robot VLA that searches over subtask sequences at test time with a world model

τ0-VLA: Robot VLA that searches over subtask sequences at test time with a world model

@jiqizhixin · 43 views · 2026-08-07

VLA世界模型World Model

Zetta ζ: Closed-Loop Recovery Harness on Frozen VLA Policies

Zetta ζ: Closed-Loop Recovery Harness on Frozen VLA Policies

@techniahqrobot · 53 views · 2026-08-30

ZettaVLA失败恢复

EchoWM: A Drivable World Model, Not Just a Clip

EchoWM: A Drivable World Model, Not Just a Clip

@JinWeiyang18434 · 20 views · 2026-08-24

世界模型World ModelEchoWM