Skip to content
Trending storyDeveloping

Liber AI releases Liber-0 robot model, 36% zero-shot success on 60 tasks

1 article1 sourcesince Nov 4Last article 1h ago ·

Overview

AISummary of 1 article

Liber AI announces Liber-0, a World Action Model for zero-shot robot manipulation, pretrained on 2.2M+ hours of human experience.

The model jointly predicts future observations and actions to turn that human experience into robot skills.

In 600 trials across 60 tasks proposed by GPT-6, Liber-0 achieved 36% overall success without robot demonstrations collected for those specific tasks. Per a report by Santiago (@svpino) on X, success was 63.5% on pick-and-place, 34.0% on other manipulation, and 10.5% on challenging multi-step tasks.

Written by AI from the articles below · updated Nov 4, 8:37 AM ET

Check the sources:

Latest developmentNov 4, 8:26 AM ET
Liber-0 learns robot manipulation from human experience, zero-shot

Article timeline

The articles in this story. Times are ET.

Nov 4
  1. SantiagoX
    Liber-0 learns robot manipulation from human experience, zero-shot

    AILiber-0, a World Action Model from Liber AI, lets robots perform zero-shot manipulation without robot demonstrations collected for each task. Across 60 tasks and 600 trials, it succeeded 63.5% of the time on pick-and-place, 34.0% on other manipulation, and 10.5% on challenging multi-step tasks, for 36% overall. The model is pretrained on 2.2M+ hours of human experience.

Heat Trend

Not enough continuous observations to show a trend yet.