# Technical Report: One-Step Drifting Action Heads for GR00T N1.7

arXiv: https://arxiv.org/abs/2609.18108 · HTML: https://mlsanity.com/paper/2609.18108/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Efficient VLA inference and adaptation, one of 13 directions in Vision-language-action models.
- First posted on arXiv 16 Sep 2026.

## Placed in

- Vision-language-action models / [Efficient VLA inference and adaptation](https://mlsanity.com/vla/efficient-vla-inference-and-adaptation/)

## Abstract (arXiv)

One-step action generation can substantially reduce the inference cost of vision-language-action (VLA) policies, but its effect on closed-loop task success remains an open question. This technical report studies a GR00T N1.7 variant in which the iterative diffusion-transformer action head is replaced by a one-step drifting action head, together with an overlap-conditioned extension for asynchronous chunk replacement. All multi-seed drifting runs were trained on two NVIDIA A800 GPUs. On LIBERO, the action head reduces the mean model-forward time of the action head from approximately 45.3,ms to 5.0,ms, while the measured backbone-plus-head time falls from approximately 70.0,ms to 30.6,ms. However, this speedup is accompanied by a systematic reduction in task success. Across three drifting seeds, success is 64.0±4.0% on LIBERO-Spatial, 52.0±1.0% on LIBERO-Goal, and 26.0±2.6% on LIBERO-Long. The low seed variance indicates that the degradation is not explained by random initialization alone. We report the result as a speed--success trade-off rather than an overall improvement, and discuss likely contributing factors including deterministic one-step mode averaging, batch-dependent geometry estimation, long open-loop chunk execution, and the fact that synchronous LIBERO evaluation does not exercise the asynchronous overlap path.