Long-term Traffic Simulation with Interleaved Autoregressive Motion and Scenario Generation
Not on the map. Why?
Abstract (arXiv)
An ideal traffic simulator replicates the realistic long-term point-to-point trip that a self-driving system experiences during deployment. Prior models and benchmarks focus on closed-loop motion simulation for initial agents in a scene. This is problematic for long-term simulation. Agents enter and exit the scene as the ego vehicle enters new regions. We propose InfGen, a unified next-token prediction model that performs interleaved closed-loop motion simulation and scene generation. InfGen automatically switches between closed-loop motion simulation and scene generation mode. It enables stable long-term rollout simulation. InfGen performs at the state-of-the-art in short-term (9s) traffic simulation, and significantly outperforms all other methods in long-term (30s) simulation. The code and model of InfGen will be released at https://orangesodahub.github.io/InfGen
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 40 of its 54 references that are in the map. First posted on arXiv 20 Jun 2025.
Builds on
40 of 54 references.
References · 40
- 2024Chameleon: Mixed-Modal Early-Fusion Foundation Modelscited by 904
- 2024Show-o: One Single Transformer to Unify Multimodal Understanding and Generationcited by 734
- 2021Large Scale Interactive Motion Forecasting for Autonomous Driving : The Waymo Open Motion Datasetcited by 532
- 2021NuPlan: A closed-loop ML-based planning benchmark for autonomous vehiclescited by 485
- 2021MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learningcited by 281
- 2022Wayformer: Motion Forecasting via Simple & Efficient Attention Networkscited by 279
- 2021TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviorscited by 192
- 2022Guided Conditional Diffusion for Controllable Traffic Simulationcited by 189
- 2023MotionLM: Multi-Agent Motion Forecasting as Language Modelingcited by 154
- 2021Generating Useful Accident-Prone Driving Scenarios via a Learned Traffic Priorcited by 136
- 2021AdvSim: Generating Safety-Critical Scenarios for Self-Driving Vehiclescited by 132
- 2019Multi-Agent Tensor Fusion for Contextual Trajectory Predictioncited by 131
28 more
- 2025Interactive Post-Training for Vision-Language-Action Modelscited by 129
- 2020Implicit Latent Variable Model for Scene-Consistent Motion Forecastingcited by 116
- 2023Language-Guided Traffic Simulation via Scene-Level Diffusioncited by 110
- 2022TrafficGen: Learning to Generate Diverse and Realistic Traffic Scenarioscited by 104
- 2023The Waymo Open Sim Agents Challengecited by 86
- 2023Language Conditioned Traffic Generationcited by 86
- 2021SceneGen: Learning to Generate Realistic Traffic Scenescited by 83
- 2022BITS: Bi-level Imitation for Traffic Simulationcited by 80
- 2024MM-Interleaved: Interleaved Image-Text Generative Modeling via Multi-modal Feature Synchronizercited by 80
- 2017Flow: A Modular Learning Framework for Mixed Autonomy Trafficcited by 67
- 2025Robust Autonomy Emerges from Self-Playcited by 64
- 2023Trajeglish: Traffic Modeling as Next-Token Predictioncited by 59
- 2024BehaviorGPT: Smart Agent Simulation for Autonomous Driving with Next-Patch Predictioncited by 49
- 2023RealGen: Retrieval Augmented Generation for Controllable Traffic Scenarioscited by 46
- 2024Orthus: Autoregressive Interleaved Image-Text Generation with Modality-Specific Headscited by 44
- 2024Solving Motion Planning Tasks with a Scalable Generative Modelcited by 40
- 2024Closed-Loop Supervised Fine-Tuning of Tokenized Traffic Modelscited by 40
- 2023SAFE-SIM: Safety-Critical Closed-Loop Traffic Simulation with Diffusion-Controllable Adversariescited by 36
- 2024SLEDGE: Synthesizing Driving Environments with Generative Models and Rule-Based Trafficcited by 27
- 2025Scenario Dreamer: Vectorized Latent Diffusion for Generating Driving Simulation Environmentscited by 27
- 2024CtRL-Sim: Reactive and Controllable Driving Agents with Offline Reinforcement Learningcited by 26
- 2023Learning Realistic Traffic Agents in Closed-loopcited by 24
- 2024Promptable Closed-loop Traffic Simulationcited by 20
- 2024AdvDiffuser: Generating Adversarial Safety-Critical Driving Scenarios via Guided Diffusioncited by 20
- 2024KiGRAS: Kinematic-Driven Generative Model for Realistic Agent Simulationcited by 17
- 2024TrafficBots V1.5: Traffic Simulation via Conditional VAEs and Transformers with Relative Pose Encodingcited by 13
- 2024Modality-Specialized Synergizers for Interleaved Vision-Language Generalistscited by 9
- 2023RMMDet: Road-Side Multitype and Multigroup Sensor Detection System for Autonomous Drivingcited by 3
Built on by
Citing papers per year: 2025: 1; 2026: 1.
2 papers.