Categories: FAANG

REFACTOR-VLA: Unsupervised Library Learning of Typed Motor Programs

Most current vision-language-action (VLA) models—such as OpenVLA, π0, RT-2, and RDT-1B—are “monolithic.” This means they generate raw motor commands or very short sequences of actions, without organizing behaviors into reusable, well-defined abstractions. As a result, these models perform poorly on long-horizon (multi-step) tasks, and it’s difficult to interpret what they have learned. Existing approaches for discovering skills often avoid the core problem of deciding when two action sequences are “behaviorally equivalent.” For example, AtomicVLA and AtomSkill group action sequences by…
AI Generated Robotic Content

Recent Posts

Accessing OpenAI models on Amazon Bedrock from Australia with global cross-Region inference

Australian teams working with OpenAI models can now access the latest OpenAI models through Amazon…

50 seconds ago

Getting started with Mantis, our open-source bug finding-and-fixing harness

AI models have clearly proven their ability to discover and exploit vulnerabilities without much, if…

52 seconds ago

Meta Pushes Its New AI Agent on Employees—but Eases Off on Tokenmaxxing

The company is reducing pressure on workers to use artificial intelligence tools while encouraging them…

1 hour ago

Why did your robotaxi stop? New system helps predict self-driving car mistakes

Self-driving cars are often controlled by deep learning models that sometimes fail in unexpected situations.…

1 hour ago

Introducing Claude Fable 5.1 on AWS

Today, we’re excited to announce the availability of Claude Fable 5.1 on Amazon Bedrock and…

1 day ago

The Range Rover Electric: Specs, Price, Availability

After long delays, JLR’s biggest gamble with its Range Rover brand is here with huge…

1 day ago