← Back to Videos

Two Ways to Speed Up Robot and Driving Policies

Loading video

FlashVLA and FlashDrive accelerate a robot-arm policy by 2.4x and an autonomous-driving policy by 4.7x for real-time control.

VLA Robot Learning Autonomous Driving Model Acceleration Real-time control Vision-language-action (VLA)

Category: research

Author: @zhijianliu

Date: 2026-11-09T00:00:00

Duration: 51.6s

Source: https://x.com/zhijianliu/status/2119853813905625504

Long-WAM: Scaling World-Action Model Context to 19.2s in Real Time

Long-WAM: Scaling World-Action Model Context to 19.2s in Real Time

@AaronWeiHuang · 593 views · 2026-10-08

World-action modelsVLA

RT-2: Turning Vision and Language Directly Into Robot Actions

RT-2: Turning Vision and Language Directly Into Robot Actions

@Alacritic_Super · 685 views · 2026-09-22

VLAVision-language-action (VLA)

Odyssey-3: A Foundation World Model for Physical Agents

Odyssey-3: A Foundation World Model for Physical Agents

@odysseyml · 1.3k views · 2026-09-15

World ModelsFoundation Models

VLA-Precision: one hour of real-robot online RL reaches 98.3% success

VLA-Precision: one hour of real-robot online RL reaches 98.3% success

@heetezition · 1.1k views · 2026-09-08

VLAReinforcement Learning

VLAct: representation-centric VLA pre-training

VLAct: representation-centric VLA pre-training

@yokoshaan · 915 views · 2026-09-05

A CoRL Keynote Teaser From Jim Fan

A CoRL Keynote Teaser From Jim Fan

@DrJimFan · 66 views · 2026-11-07

Robot LearningEmbodied AI

As an Amazon Associate, we earn from qualifying purchases.