mlsanityData to 7 Oct 2026

Are VLMs Ready for Autonomous Driving? An Empirical Study from the Reliability, Data, and Metric Perspectives

arXiv 2501.04003 ↗First posted 7 Jan 2025cs.CV cs.ROcited by 120

On the map: Multimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection

also in 1 more
Abstract (arXiv)

Recent advancements in Vision-Language Models (VLMs) have sparked interest in their use for autonomous driving, particularly in generating interpretable driving decisions through natural language. However, the assumption that VLMs inherently provide visually grounded, reliable, and interpretable explanations for driving remains largely unexamined. To address this gap, we introduce DriveBench, a benchmark dataset designed to evaluate VLM reliability across 17 settings (clean, corrupted, and text-only inputs), encompassing 19,200 frames, 20,498 question-answer pairs, three question types, four mainstream driving tasks, and a total of 12 popular VLMs. Our findings reveal that VLMs often generate plausible responses derived from general knowledge or textual cues rather than true visual grounding, especially under degraded or missing visual inputs. This behavior, concealed by dataset imbalances and insufficient evaluation metrics, poses significant risks in safety-critical scenarios like autonomous driving. We further observe that VLMs struggle with multi-modal reasoning and display heightened sensitivity to input corruptions, leading to inconsistencies in performance. To address these challenges, we propose refined evaluation metrics that prioritize robust visual grounding and multi-modal understanding. Additionally, we highlight the potential of leveraging VLMs' awareness of corruptions to enhance their reliability, offering a roadmap for developing more trustworthy and interpretable decision-making systems in real-world autonomous driving contexts. The benchmark toolkit is publicly accessible.

Key facts. Cited by 120 papers in the map (data through 7 Oct 2026). Placed in Driving VLMs and open-vocabulary detection, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 65 of its 89 references that are in the map. First posted on arXiv 7 Jan 2025.

Builds on

65 of 89 references.

Closest ancestors · 8

Same direction, newest first

  1. 2024VLM-AD: End-to-End Autonomous Driving through Vision-Language Model Supervision19 Dec
  2. 2024Senna: Bridging Large Vision-Language Models and End-to-End Autonomous Driving29 Oct
  3. 2024V2X-VLM: End-to-End V2X Cooperative Autonomous Driving Through Large Vision-Language Models17 Aug
  4. 2024Automated Evaluation of Large Vision-Language Models on Self-driving Corner Cases16 Apr
  5. 2024DriveVLM: The Convergence of Autonomous Driving and Large Vision-Language Models19 Feb
  6. 2023DriveLM: Driving with Graph Visual Question Answering21 Dec
  7. 2023Dolphins: Multimodal Language Model for Driving1 Dec
  8. 2023DriveGPT4: Interpretable End-to-end Autonomous Driving via Large Language Model2 Oct

Other references · 57

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2023Visual Instruction Tuningcited by 7,792
  3. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  4. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  5. 2023Improved Baselines with Visual Instruction Tuningcited by 4,362
  6. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,151
  7. 2023Qwen Technical Reportcited by 3,273
  8. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  9. 2023InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Taskscited by 2,534
  10. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,436
  11. 2022Survey of Hallucination in Natural Language Generationcited by 2,056
  12. 2014CIDEr: Consensus-based Image Description Evaluationcited by 2,027
38 more
  1. 2024Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phonecited by 1,883
  2. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,696
  3. 2018BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learningcited by 1,282
  4. 2022BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformerscited by 1,213
  5. 2022Planning-oriented Autonomous Drivingcited by 1,068
  6. 2024How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suitescited by 1,033
  7. 2016SPICE: Semantic Propositional Image Caption Evaluationcited by 820
  8. 2021BEVDet: High-performance Multi-camera 3D Object Detection in Bird-Eye-Viewcited by 711
  9. 2023CogAgent: A Visual Language Model for GUI Agentscited by 692
  10. 2021DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queriescited by 655
  11. 2023LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Modelcited by 602
  12. 2024Justice or Prejudice? Quantifying Biases in LLM-as-a-Judgecited by 354
  13. 2023GPT-Driver: Learning to Drive with GPTcited by 344
  14. 2024Humans or LLMs as the Judge? A Study on Judgement Biasescited by 314
  15. 2023NuScenes-QA: A Multi-modal Visual Question Answering Benchmark for Autonomous Driving Scenariocited by 258
  16. 2023Is Ego Status All You Need for Open-Loop End-to-End Autonomous Driving?cited by 242
  17. 2023Driving with LLMs: Fusing Object-Level Vector Modality for Explainable Autonomous Drivingcited by 231
  18. 2018Textual Explanations for Self-Driving Vehiclescited by 226
  19. 2023DiLu: A Knowledge-Driven Approach to Autonomous Driving with Large Language Modelscited by 204
  20. 2023Drive Like a Human: Rethinking Autonomous Driving with Large Language Modelscited by 158
  21. 2024Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolutioncited by 157
  22. 2023Open-World Object Manipulation using Pre-trained Vision-Language Modelscited by 155
  23. 2024Aligning with Human Judgement: The Role of Pairwise Preference in Large Language Model Evaluatorscited by 146
  24. 2019Talk2Car: Taking Control of Your Self-Driving Carcited by 141
  25. 2022Data Determines Distributional Robustness in Contrastive Language Image Pre-training (CLIP)cited by 137
  26. 2023LLM4Drive: A Survey of Large Language Models for Autonomous Drivingcited by 136
  27. 2023Robo3D: Towards Robust and Reliable 3D Perception against Corruptionscited by 128
  28. 2022DRAMA: Joint Risk Localization and Captioning in Drivingcited by 125
  29. 2024VLP: Vision Language Planning for Autonomous Drivingcited by 125
  30. 2024Insight-V: Exploring Long-Chain Visual Reasoning with Multimodal Large Language Modelscited by 112
  31. 2023Language Prompt for Autonomous Drivingcited by 110
  32. 2023How Many Unicorns Are in This Image? A Safety Evaluation Benchmark for Vision LLMscited by 103
  33. 2023Drive as You Speak: Enabling Human-Like Interaction with Large Language Models in Autonomous Vehiclescited by 100
  34. 2024DrivingGPT: Unifying Driving World Modeling and Planning with Multi-modal Autoregressive Transformerscited by 96
  35. 2020Explainable Object-induced Action Decision for Autonomous Vehiclescited by 88
  36. 2023Octopus: Embodied Vision-Language Programmer from Environmental Feedbackcited by 73
  37. 2022Out of the BLEU: how should we assess quality of the Code Generation models?cited by 65
  38. 2024Chain-of-Spot: Interactive Reasoning Improves Large Vision-Language Modelscited by 65

50 most cited of 57 All 57 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 61; 2026: 59.

120 papers.

Sort:

By area:

  1. 2025OpenDriveVLA: Towards End-to-end Autonomous Driving with Large Vision Language Action Modelcited by 172
  2. 2025Alpamayo-R1: Bridging Reasoning and Action Prediction for Generalizable Autonomous Driving in the Long Tailcited by 136
  3. 2025OmniSpatial: Towards Comprehensive Spatial Reasoning Benchmark for Vision Language Modelscited by 105
  4. 2025A Survey on Vision-Language-Action Models for Autonomous Drivingcited by 72
  5. 2025Impromptu VLA: Open Weights and Open Data for Driving Vision-Language-Action Modelscited by 60
  6. 2025How Far are VLMs from Visual Spatial Intelligence? A Benchmark-Driven Perspectivecited by 41
  7. 2025Foundation Models in Autonomous Driving: A Survey on Scenario Generation and Scenario Analysiscited by 37
  8. 2025AgentThink: A Unified Framework for Tool-Augmented Chain-of-Thought Reasoning in Vision-Language Models for Autonomous Drivingcited by 35
  9. 2025Vision-Language-Action Models for Autonomous Driving: Past, Present, and Futurecited by 32
  10. 2025PixelThink: Towards Efficient Chain-of-Pixel Reasoningcited by 28
  11. 2025Exploring the Roles of Large Language Models in Reshaping Transportation Systems: A Survey, Framework, and Roadmapcited by 26
  12. 2025Fine-Grained Evaluation of Large Vision-Language Models in Autonomous Drivingcited by 19
38 more
  1. 2026VGGDrive: Empowering Vision-Language Models with Cross-View Geometric Grounding for Autonomous Drivingcited by 19
  2. 2025DriveAction: A Benchmark for Exploring Human-like Driving Decisions in VLA Modelscited by 17
  3. 2025STSBench: A Spatio-temporal Scenario Benchmark for Multi-modal Large Language Models in Autonomous Drivingcited by 17
  4. 2026UniDriveVLA: Unifying Understanding, Perception, and Action Planning for Autonomous Drivingcited by 17
  5. 2025Perspective-Invariant 3D Object Detectioncited by 16
  6. 2025Reinforcing Video Reasoning Segmentation to Think Before It Segmentscited by 15
  7. 2025WAM-Flow: Parallel Coarse-to-Fine Motion Planning via Discrete Flow Matching for Autonomous Drivingcited by 15
  8. 2025LiDARCrafter: Dynamic 4D World Modeling from LiDAR Sequencescited by 14
  9. 2025Multi-Crit: Benchmarking Multimodal Judges on Pluralistic Criteria-Followingcited by 14
  10. 2025VDT-Auto: End-to-end Autonomous Driving with VLM-Guided Diffusion Transformerscited by 12
  11. 2025Chain-of-Thought for Autonomous Driving: A Comprehensive Survey and Future Prospectscited by 12
  12. 2025SpinBench: Perspective and Rotation as a Lens on Spatial Reasoning in VLMscited by 12
  13. 2025Extending Large Vision-Language Model for Diverse Interactive Tasks in Autonomous Drivingcited by 10
  14. 2025DriveQA: Passing the Driving Knowledge Testcited by 10
  15. 2025Actial: Activate Spatial Reasoning Ability of Multimodal Large Language Modelscited by 10
  16. 2025CHAI: Command Hijacking against embodied AIcited by 9
  17. 2025WaymoQA: A Multi-View Visual Question Answering Dataset for Safety-Critical Reasoning in Autonomous Drivingcited by 9
  18. 2026PhyCritic: Multimodal Critic Models for Physical AIcited by 9
  19. 2025Bench2ADVLM: A Closed-Loop Benchmark for Vision-language Models in Autonomous Drivingcited by 8
  20. 2025AccidentBench: Benchmarking Multimodal Understanding and Reasoning in Vehicle Accidents and Beyondcited by 8
  21. 2025RoboDriveVLM: A Novel Benchmark and Baseline towards Robust Vision-Language Models for Autonomous Drivingcited by 8
  22. 2026Bench2Drive-VL: Benchmarks for Closed-Loop Autonomous Driving with Vision-Language Modelscited by 8
  23. 2026nuReasoning: A Reasoning-Centric Dataset and Benchmark for Long-Tail Autonomous Drivingcited by 8
  24. 2025Generative AI for Autonomous Driving: Frontiers and Opportunitiescited by 7
  25. 2025Beyond One Shot, Beyond One Perspective: Cross-View and Long-Horizon Distillation for Better LiDAR Representationscited by 7
  26. 2025CoReVLA: A Dual-Stage End-to-End Autonomous Driving Framework for Long-Tail Scenarios via Collect-and-Refinecited by 7
  27. 2026CausalVAD: De-confounding End-to-End Autonomous Driving via Causal Interventioncited by 7
  28. 2025Talk2Event: Grounded Understanding of Dynamic Scenes from Event Camerascited by 6
  29. 2025GraphPilot: Grounded Scene Graph Conditioning for Language-Based Autonomous Drivingcited by 6
  30. 2025DriveRX: A Vision-Language Reasoning Model for Cross-Task Autonomous Drivingcited by 5
  31. 2025VDRive: Leveraging Reinforced VLA and Diffusion Policy for End-to-end Autonomous Drivingcited by 5
  32. 2025Evaluation of Safety Cognition Capability in Vision-Language Models for Autonomous Drivingcited by 4
  33. 2025VLAD: A VLM-Augmented Autonomous Driving Framework with Hierarchical Planning and Interpretable Decision Processcited by 4
  34. 2025LMAD: Integrated End-to-End Vision-Language Model for Explainable Autonomous Drivingcited by 4
  35. 2025Forging Spatial Intelligence: A Roadmap of Multi-Modal Data Pre-Training for Autonomous Systemscited by 4
  36. 2026AgentDrive: An Open Benchmark Dataset for Agentic AI Reasoning with LLM-Generated Scenarios in Autonomous Systemscited by 4
  37. 2025Structured Labeling Enables Faster Vision-Language Models for End-to-End Autonomous Drivingcited by 3
  38. 2025More Than Meets the Eye? Uncovering the Reasoning-Planning Disconnect in Training Vision-Language Driving Modelscited by 3

50 most cited of 120 All 120 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Probing the Reliability of Driving VLMs: From Inconsistent Responses to Grounded Temporal Reasoning
  2. 2025VLMs Guided Interpretable Decision Making for Autonomous Driving
  3. 2025LightEMMA: A Longitudinal Evaluation of Vision-Language Models for Autonomous Driving
  4. 2023Vision Language Models in Autonomous Driving: A Survey and Outlook
  5. 2025Are Vision LLMs Road-Ready? A Comprehensive Benchmark for Safety-Critical Driving Video Understanding
  6. 2025A Hierarchical Test Platform for Vision Language Model (VLM)-Integrated Real-World Autonomous Driving
  7. 2025ReasonDrive: Efficient Visual Question Answering for Autonomous Vehicles with Reasoning-Enhanced Small Vision-Language Models
  8. 2025AutoDrive-QA: A Multiple-Choice Benchmark for Vision-Language Evaluation in Urban Autonomous Driving