# Judge, Then Drive: A Critic-Centric Vision Language Action Framework for Autonomous Driving

arXiv: https://arxiv.org/abs/2604.27366 · HTML: https://mlsanity.com/paper/2604.27366/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in VLA for autonomous driving, one of 13 directions in Vision-language-action models.
- Builds on 30 of its 69 references that are in the map.
- First posted on arXiv 30 Apr 2026.

## Placed in

- Vision-language-action models / [VLA for autonomous driving](https://mlsanity.com/vla/vla-for-autonomous-driving/)
- End-to-end autonomous driving / [Reasoning VLA and VLM driving policies](https://mlsanity.com/driving/reasoning-vla-and-vlm-driving-policies/)

## Builds on (most cited first, 30 of 30)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [OpenVLA: An Open-Source Vision-Language-Action Model](https://mlsanity.com/paper/2406.09246/) (2024)
- [Qwen2 Technical Report](https://mlsanity.com/paper/2407.10671/) (2024)
- [Decoupled Weight Decay Regularization](https://mlsanity.com/paper/1711.05101/) (2017)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate](https://mlsanity.com/paper/2305.19118/) (2023)
- [Argoverse 2: Next Generation Datasets for Self-Driving Perception and Forecasting](https://mlsanity.com/paper/2301.00493/) (2023)
- [RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback](https://mlsanity.com/paper/2309.00267/) (2023)
- [A Survey on Vision-Language-Action Models for Embodied AI](https://mlsanity.com/paper/2405.14093/) (2024)
- [Auto-GPT for Online Decision Making: Benchmarks and Additional Opinions](https://mlsanity.com/paper/2306.02224/) (2023)
- [Bench2Drive: Towards Multi-Ability Benchmarking of Closed-Loop End-To-End Autonomous Driving](https://mlsanity.com/paper/2406.03877/) (2024)
- [ORION: A Holistic End-to-End Autonomous Driving Framework by Vision-Language Instructed Action Generation](https://mlsanity.com/paper/2503.19755/) (2025)
- [DriveTransformer: Unified Transformer for Scalable End-to-End Autonomous Driving](https://mlsanity.com/paper/2503.07656/) (2025)
- [PRD: Peer Rank and Discussion Improve Large Language Model based Evaluations](https://mlsanity.com/paper/2307.02762/) (2023)
- [RAG-Driver: Generalisable Driving Explanations with Retrieval-Augmented In-Context Learning in Multi-Modal Large Language Model](https://mlsanity.com/paper/2402.10828/) (2024)
- [Vision-Language-Action (VLA) Models: Concepts, Progress, Applications and Challenges](https://mlsanity.com/paper/2505.04769/) (2025)
- [A Survey on Vision-Language-Action Models for Autonomous Driving](https://mlsanity.com/paper/2506.24044/) (2025)
- [DriveCoT: Integrating Chain-of-Thought Reasoning with End-to-End Driving](https://mlsanity.com/paper/2403.16996/) (2024)
- [DiffVLA: Vision-Language Guided Diffusion Planning for Autonomous Driving](https://mlsanity.com/paper/2505.19381/) (2025)
- [RouterEval: A Comprehensive Benchmark for Routing LLMs to Explore Model-level Scaling Up in LLMs](https://mlsanity.com/paper/2503.10657/) (2025)
- [Human-Centered Design Recommendations for LLM-as-a-Judge](https://mlsanity.com/paper/2407.03479/) (2024)
- [LLaVA-Critic-R1: Your Critic Model is Secretly a Strong Policy Model](https://mlsanity.com/paper/2509.00676/) (2025)
- [Hidden Biases of End-to-End Driving Datasets](https://mlsanity.com/paper/2412.09602/) (2024)
- [SafeAuto: Knowledge-Enhanced Safe Autonomous Driving with Multimodal Foundation Models](https://mlsanity.com/paper/2503.00211/) (2025)
- [BridgeDrive: Diffusion Bridge Policy for Closed-Loop Trajectory Planning in Autonomous Driving](https://mlsanity.com/paper/2509.23589/) (2025)
- [A Closer Look into Automatic Evaluation Using Large Language Models](https://mlsanity.com/paper/2310.05657/) (2023)
- [Chain-of-Thought for Autonomous Driving: A Comprehensive Survey and Future Prospects](https://mlsanity.com/paper/2505.20223/) (2025)
- [MiniLongBench: The Low-cost Long Context Understanding Benchmark for Large Language Models](https://mlsanity.com/paper/2505.19959/) (2025)
- [HeadCraft: Modeling High-Detail Shape Variations for Animated 3DMMs](https://mlsanity.com/paper/2312.14140/) (2023)

## Built on by (most cited first, 3 of 3)

- [BLUE: Toward Better Language Use in Efficient Vision-Language-Action Models for Autonomous Driving](https://mlsanity.com/paper/2606.08684/) (2026)
- [Silent Failures in Physical AI: A Literature Review of Runtime Action Authorization for Autonomous Systems](https://mlsanity.com/paper/2606.00090/) (2026)
- [Drive by Hindsight and Foresight: Tool-Grounded Synergistic Reasoning over Hierarchical Memory for Autonomous Driving](https://mlsanity.com/paper/2609.08217/) (2026)

## Abstract (arXiv)

Recent advances in vision language action (VLA) models have shown remarkable potential for autonomous driving by directly mapping multimodal inputs to control signals. However, previous VLA-based methods have not explicitly exploited the critic capability of VLAs to refine driving decisions, even though such capability has been well demonstrated in other LLM-based domains, thereby limiting their performance in complex closed-loop scenarios. In this work, we present a theoretically inspired two-stage framework, CriticVLA, which extends the role of VLAs from acting to judging. CriticVLA first generates a rough trajectory and then refines it through multimodal evaluation and single-step optimization guided by a VLA-based critic, yielding higher-quality driving behaviors. To support this process, we construct a large-scale synthetic dataset of 12.9 million annotated trajectories covering diverse driving scenarios, which enhances the critic's reasoning and refinement abilities. Extensive closed-loop experiments on the Bench2Drive benchmark show that CriticVLA significantly surpasses state-of-the-art baselines, achieving a 73.33% total success rate and delivering about 30% improvement in challenging scenarios.