NinA: Normalizing Flows in Action. Training VLA Models with Normalizing Flows

TL;DR

NinA introduces a Normalizing Flow-based decoder for Vision-Language-Action models, enabling faster, one-shot action sampling while maintaining performance comparable to diffusion models, thus improving real-time control capabilities.

Contribution

The paper proposes NinA, a novel Normalizing Flow-based decoder for VLA models, replacing diffusion models to enable rapid, one-shot action sampling without performance loss.

Findings

NinA matches diffusion models in performance on LIBERO benchmark.

NinA significantly reduces inference time for VLA models.

NinA enables high-frequency control suitable for real-world applications.

Abstract

Recent advances in Vision-Language-Action (VLA) models have established a two-component architecture, where a pre-trained Vision-Language Model (VLM) encodes visual observations and task descriptions, and an action decoder maps these representations to continuous actions. Diffusion models have been widely adopted as action decoders due to their ability to model complex, multimodal action distributions. However, they require multiple iterative denoising steps at inference time or downstream techniques to speed up sampling, limiting their practicality in real-world settings where high-frequency control is crucial. In this work, we present NinA (Normalizing Flows in Action), a fast and expressive alternative to diffusion-based decoders for VLAs. NinA replaces the diffusion action decoder with a Normalizing Flow (NF) that enables one-shot sampling through an invertible transformation, significantly reducing inference time. We integrate NinA into the FLOWER VLA architecture and fine-tune on the LIBERO benchmark. Our experiments show that NinA matches the performance of its diffusion-based counterpart under the same training regime, while achieving substantially faster inference. These results suggest that NinA offers a promising path toward efficient, high-frequency VLA control without compromising performance.