NoRD: A Data-Efficient Vision-Language-Action Model that Drives without Reasoning

2026-02-24

AI Summary

This work proposes NoRD, a novel vision-language-action (VLA) driving model that achieves efficient end-to-end policy learning without relying on expensive reasoning annotations. Addressing the high training costs of existing VLA approaches—which depend on large-scale datasets and densely annotated reasoning traces—NoRD operates effectively on significantly smaller datasets (over 40% reduction in data volume). The method introduces Dr. GRPO, an optimized variant of the GRPO algorithm designed to mitigate optimization failure and difficulty bias commonly observed in small-data regimes. Evaluated on the Waymo and NAVSIM benchmarks, NoRD matches or approaches the performance of current state-of-the-art methods while substantially improving training efficiency and eliminating the need for costly reasoning labels.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Vision-Language-Action (VLA) models are advancing autonomous driving by replacing modular pipelines with unified end-to-end architectures. However, current VLAs face two expensive requirements: (1) massive dataset collection, and (2) dense reasoning annotations. In this work, we address both challenges with \modelname (\textbf{No} \textbf{R}easoning for \textbf{D}riving). Compared to existing VLAs, \modelname achieves competitive performance while being fine-tuned on $<$60\% of the data and no reasoning annotations, resulting in 3$\times$ fewer tokens. We identify that standard Group Relative Policy Optimization (GRPO) fails to yield significant improvements when applied to policies trained on such small, reasoning-free datasets. We show that this limitation stems from difficulty bias, which disproportionately penalizes reward signals from scenarios that produce high-variance rollouts within GRPO. \modelname overcomes this by incorporating Dr.~GRPO, a recent algorithm designed to mitigate difficulty bias in LLMs. As a result, \modelname achieves competitive performance on Waymo and NAVSIM with a fraction of the training data and no reasoning overhead, enabling more efficient autonomous systems.

Problem

• Vision-Language-Action
• autonomous driving
• data efficiency
• reasoning annotations
• difficulty bias

Innovation

• Vision-Language-Action
• data-efficient learning
• reasoning-free training
• Dr. GRPO
• difficulty bias