Open Models Almanac
Methods

single mention · not yet filed

L1 regression objective for continuous actions

L1 regression is a method used for continuous actions in robot policies.1 It enables single-pass action generation and effectively models the multi-task action distribution.2,3

Trade-off The simpler algorithm leads to faster training convergence and inference speed while maintaining strong performance.4

How this definition was written

Summarised by glm-5.3-flash from 1 source; every sentence is checked against the quote it cites.

  1. continuous actions with L1 regression — Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
  2. L1 regression-based single-pass action generation — Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
  3. L1 regression and diffusion variants achieve comparable performance, indicating that the high-capacity OpenVLA model can effectively model the multi-task action distribution even with simple L1 regression — Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
  4. the simpler algorithm leads to faster training convergence and inference speed while maintaining strong performance — Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success

Also called L1 regression.