mlsanityData to 14 Oct 2026

Raw2Drive: Reinforcement Learning with Aligned World Models for End-to-End Autonomous Driving (in CARLA v2)

arXiv 2505.16394 ↗First posted 22 May 2025cs.RO cs.AI cs.CVcited by 47

On the map: World models → World models for autonomous driving

also in 3 more
Abstract (arXiv)

Reinforcement Learning (RL) can mitigate the causal confusion and distribution shift inherent to imitation learning (IL). However, applying RL to end-to-end autonomous driving (E2E-AD) remains an open problem for its training difficulty, and IL is still the mainstream paradigm in both academia and industry. Recently Model-based Reinforcement Learning (MBRL) have demonstrated promising results in neural planning; however, these methods typically require privileged information as input rather than raw sensor data. We fill this gap by designing Raw2Drive, a dual-stream MBRL approach. Initially, we efficiently train an auxiliary privileged world model paired with a neural planner that uses privileged information as input. Subsequently, we introduce a raw sensor world model trained via our proposed Guidance Mechanism, which ensures consistency between the raw sensor world model and the privileged world model during rollouts. Finally, the raw sensor world model combines the prior knowledge embedded in the heads of the privileged world model to effectively guide the training of the raw sensor policy. Raw2Drive is so far the only RL based end-to-end method on CARLA Leaderboard 2.0, and Bench2Drive and it achieves state-of-the-art performance.

Key facts. Cited by 47 papers in the map (data through 14 Oct 2026). Placed in World models for autonomous driving, one of 11 directions in World models. Builds on 48 of its 72 references that are in the map. First posted on arXiv 22 May 2025.

Builds on

48 of 72 references.

Closest ancestors · 6

Same direction, newest first

  1. 2024Enhancing End-to-End Autonomous Driving with Latent World Model12 Jun
  2. 2024Think2Drive: Efficient Reinforcement Learning by Thinking in Latent World Model for Quasi-Realistic Autonomous Driving (in CARLA-v2)26 Feb
  3. 2024GenAD: Generative End-to-End Autonomous Driving18 Feb
  4. 2022BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers31 Mar
  5. 2019Model-free Deep Reinforcement Learning for Urban Autonomous Driving20 Apr
  6. 2017CARLA: An Open Urban Driving Simulator10 Nov

Other references · 42

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  3. 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
  4. 2023Mastering Diverse Domains through World Modelscited by 1,253
  5. 2022Planning-oriented Autonomous Drivingcited by 1,087
  6. 2024WorldSimBench: Towards Video Generation Models as World Simulatorscited by 881
  7. 2017Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithmcited by 734
  8. 2023VAD: Vectorized Scene Representation for Efficient Autonomous Drivingcited by 622
  9. 2023End-to-end Autonomous Driving: Challenges and Frontierscited by 490
  10. 2022TransFuser: Imitation with Transformer-Based Sensor Fusion for Autonomous Drivingcited by 485
  11. 2019Learning by Cheatingcited by 383
  12. 2021Multi-Modal Fusion Transformer for End-to-End Autonomous Drivingcited by 377
30 more
  1. 2024Bench2Drive: Towards Multi-Ability Benchmarking of Closed-Loop End-To-End Autonomous Drivingcited by 252
  2. 2022Trajectory-guided Control Prediction for End-to-end Autonomous Driving: A Simple yet Strong Baselinecited by 248
  3. 2022Safety-Enhanced Autonomous Driving Using Interpretable Sensor Fusion Transformercited by 212
  4. 2021End-to-End Urban Driving by Imitating a Reinforcement Learning Coachcited by 194
  5. 2023Rethinking the Open-Loop Evaluation of End-to-End Autonomous Driving in nuScenescited by 191
  6. 2023Think Twice before Driving: Towards Scalable Decoders for End-to-End Autonomous Drivingcited by 169
  7. 2019End-to-End Model-Free Reinforcement Learning for Urban Driving using Implicit Affordancescited by 149
  8. 2023DriveAdapter: Breaking the Coupling Barrier of Perception and Planning in End-to-End Autonomous Drivingcited by 144
  9. 2023LLM4Drive: A Survey of Large Language Models for Autonomous Drivingcited by 136
  10. 2025DriveTransformer: Unified Transformer for Scalable End-to-End Autonomous Drivingcited by 126
  11. 2022Delving into the Devils of Bird's-eye-view Perception: A Review, Evaluation and Recipecited by 122
  12. 2024RGBD Objects in the Wild: Scaling Real-World 3D Object Learning from RGB-D Videoscited by 115
  13. 2022HDGT: Heterogeneous Driving Graph Transformer for Multi-Agent Trajectory Prediction via Scene Encodingcited by 110
  14. 2023ReasonNet: End-to-End Driving with Temporal and Global Reasoningcited by 93
  15. 2025Don't Shake the Wheel: Momentum-Aware Planning in End-to-End Autonomous Drivingcited by 84
  16. 2020Fighting Copycat Agents in Behavioral Cloning from Observation Historiescited by 71
  17. 2018Probabilistic Recurrent State-Space Modelscited by 64
  18. 2025CaRL: Learning Scalable Planning Policies with Simple Rewardscited by 35
  19. 2024Hidden Biases of End-to-End Driving Datasetscited by 30
  20. 2025Amulet: ReAlignment During Test Time for Personalized Preference Adaptation of LLMscited by 29
  21. 2023LightZero: A Unified Benchmark for Monte Carlo Tree Search in General Sequential Decision Scenarioscited by 20
  22. 2020IDE-Net: Interactive Driving Event and Pattern Extraction from Human Datacited by 19
  23. 2024ActiveAD: Planning-Oriented Active Learning for End-to-End Autonomous Drivingcited by 16
  24. 2024RAT: Adversarial Attacks on Deep Reinforcement Agents for Targeted Behaviorscited by 16
  25. 2024Bench2Drive-R: Turning Real World Data into Reactive Closed-Loop Autonomous Driving Benchmark by Generative Modelcited by 14
  26. 2023Normalization Enhances Generalization in Visual Reinforcement Learningcited by 13
  27. 2024AMP: Autoregressive Motion Prediction Revisited with Next Token Prediction for Autonomous Drivingcited by 13
  28. 2024UniZero: Generalized and Efficient Planning with Scalable Latent World Modelscited by 12
  29. 2024Efficient Preference-based Reinforcement Learning via Aligned Experience Estimationcited by 9
  30. 2024FlatFusion: Delving into Details of Sparse Transformer-based Camera-LiDAR Fusion for Autonomous Drivingcited by 9

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 12; 2026: 35.

47 papers.

Sort:

By area:

  1. 2025A Comprehensive Survey on World Models for Embodied AIcited by 68
  2. 2025ReSim: Reliable World Simulation for Autonomous Drivingcited by 44
  3. 2026MIND: Benchmarking Memory Consistency and Action Control in World Modelscited by 35
  4. 2025Vision-Language-Action Models for Autonomous Driving: Past, Present, and Futurecited by 33
  5. 2025SpaceDrive: Infusing Spatial Awareness into VLM-based Autonomous Drivingcited by 28
  6. 2025MindDrive: A Vision-Language-Action Model for Autonomous Driving via Online Reinforcement Learningcited by 25
  7. 2026DriveVA: Video Action Models are Zero-Shot Driverscited by 23
  8. 2026PALM: Progress-Aware Policy Learning via Affordance Reasoning for Long-Horizon Robotic Manipulationcited by 21
  9. 2025Zero-Human Demonstration End-to-end Autonomous Driving with Trajectory Scorercited by 16
  10. 2026EponaV2: Driving World Model with Comprehensive Future Reasoningcited by 15
  11. 2025AD-R1: Closed-Loop Reinforcement Learning for End-to-End Autonomous Driving with Impartial World Modelscited by 14
  12. 2025MindDrive: An All-in-One Framework Bridging World Models and Vision-Language Model for End-to-End Autonomous Drivingcited by 12
35 more
  1. 2026DynVLA: Learning World Dynamics for Action Reasoning in Autonomous Drivingcited by 11
  2. 2026DriveMamba: Task-Centric Scalable State Space Model for Efficient End-to-End Autonomous Drivingcited by 9
  3. 2026AutoMoT: A Unified Vision-Language-Action Model with Asynchronous Mixture-of-Transformers for End-to-End Autonomous Drivingcited by 9
  4. 2026SimWAM: A Simple World Action Model for End-to-End Autonomous Drivingcited by 9
  5. 2025PlanT 2.0: Exposing Biases and Structural Flaws in Closed-Loop Drivingcited by 8
  6. 2026Learning from Mistakes: Post-Training for Driving VLA with Takeover Datacited by 8
  7. 2026World Engine: Towards the Era of Post-Training for Autonomous Drivingcited by 8
  8. 2026GraphWorld: Long-Horizon Planning with World Models for End-to-End Autonomous Drivingcited by 6
  9. 2026HAD: Combining Hierarchical Diffusion with Metric-Decoupled RL for End-to-End Drivingcited by 4
  10. 2025Large Language Model-assisted Autonomous Vehicle Recovery from Immobilizationcited by 3
  11. 2026Latent World Models for Automated Driving: A Unified Taxonomy, Evaluation Framework, and Open Challengescited by 3
  12. 2026Bench2Drive-Robust: Benchmarking Closed-Loop Autonomous Driving under Deployment Perturbationscited by 3
  13. 2026Distill to Think, Foresee to Act: Cognitive-Physical Reinforcement Learning for Autonomous Drivingcited by 3
  14. 2026DriveZero: End-to-End Driving Beyond Human Demonstrationscited by 3
  15. 2025Driving Beyond Privilege: Distilling Dense-Reward Knowledge into Sparse-Reward Policiescited by 2
  16. 2026CRAFT: Counterfactual-to-Interactive Reinforcement Fine-Tuning for Driving Policiescited by 2
  17. 2026CLEAR: Closed-Loop Reinforcement Learning at Scale for End-to-End Autonomous Drivingcited by 2
  18. 2026PerlAD: Towards Enhanced Closed-loop End-to-end Autonomous Driving with Pseudo-simulation-based Reinforcement Learningcited by 1
  19. 2026Causality-Aware End-to-End Autonomous Driving via Ego-Centric Joint Scene Modelingcited by 1
  20. 2026BLUE: Toward Better Language Use in Efficient Vision-Language-Action Models for Autonomous Drivingcited by 1
  21. 2026SimVTA: Simulation-Grounded Video-Trajectory Alignment for Driving World Action Modelscited by 1
  22. 2025Calibrating Biased Distribution in VFM-derived Latent Space via Cross-Domain Geometric Consistencycited by 0
  23. 2026Kinematics-Aware Latent World Models for Data-Efficient Autonomous Drivingcited by 0
  24. 2026Dreaming Across Towns: Semantic Rollout and Town-Adversarial Regularization for Zero-Shot Held-Out-Town Fixed-Route Driving in CARLAcited by 0
  25. 2026GraphBEV++: Multi-Modal Feature Alignment for Autonomous Drivingcited by 0
  26. 2026PlanRL: A Trajectory Planning Architecture for Reinforcement Learning-based Driving Expertscited by 0
  27. 2026Collaborative Weighting with Pessimistic Critic for Mitigating Overestimation in Off-Policy Reinforcement Learningcited by 0
  28. 2026Dreamer-SAC: Off-Policy Learning in Latent World Models for Sample-Efficient Autonomous Drivingcited by 0
  29. 2026Threat-guided Policy-aware Scene Perturbation for Safe Autonomous Driving with Online Reinforcement Learningcited by 0
  30. 2026MomADv2: Reliable Temporal Memory for End-to-End Autonomous Drivingcited by 0
  31. 2026Driving on Memorycited by 0
  32. 2026FIVE-VLA: Fast and EffectIVE Autonomous Driving with Recurrent Action Memorycited by 0
  33. 2026RoXDrive: Closed-Loop Reinforcement Learning for End-to-End Autonomous Driving via Action-Faithful Rolloutscited by 0
  34. 2026A Survey on End-to-End Autonomous Driving Training from the Perspectives of Data, Strategy, and Platformcited by 0
  35. 2026DriveWLA: Unifying Future Dynamics and Scene Semantics through World-Language-Action Modeling for Autonomous Drivingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Driving
  2. 2024From Imitation to Exploration: End-to-end Autonomous Driving based on World Model
  3. 2024Imagine-2-Drive: Leveraging High-Fidelity World Models via Multi-Modal Diffusion Policies
  4. 2020Guided Policy Search Model-based Reinforcement Learning for Urban Autonomous Driving
  5. 2020Learning to Drive (L2D) as a Low-Cost Benchmark for Real-World Reinforcement Learning
  6. 2023Privileged to Predicted: Towards Sensorimotor Reinforcement Learning for Urban Driving
  7. 2026ExploreVLA: Dense World Modeling and Exploration for End-to-End Autonomous Driving
  8. 2025Pseudo-Expert Regularized Offline RL for End-to-End Autonomous Driving in Photorealistic Closed-Loop Environments