Yann LeCun's AD-E2E-JEPA Plans Driving Trajectories 100x Faster Without Policy Training

A new JEPA world model drives without a trained policy, hitting goals 20 meters out with a 100x inference speedup over prior latent planners.

Source

AlphaSignal

Published

Sep 28, 2026, 1:22 AM

Author

AlphaSignal Newsroom

Read

1 min read

A new JEPA world model drives without a trained policy, hitting goals 20 meters out with a 100x inference speedup over prior latent planners.

Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.

Next reads