arXiv:2606. 31844v1 Announce Type: cross Abstract: A local-to-global context mismatch arises when autoregressive traffic simulators trained on ego-centric driving logs are deployed in globally observable closed-loop environments.
arXiv:2608.22187v1 Announce Type: cross Abstract: Modern driving action models are increasingly improved in a self-improvement loop, where a learned world simulator imagines future observations and t...
arXiv:2606. 31209v1 Announce Type: new Abstract: Interactive traffic simulation is a vital world model for autonomous driving.
arXiv:2609.26792v1 Announce Type: cross Abstract: Faithfully evaluating end-to-end driving policies in simulation requires observations that are not merely photo-realistic, but preserve the scene fea...
arXiv:2606. 17362v1 Announce Type: cross Abstract: Autonomous driving has shifted towards end-to-end policy learning, where reliable, interpretable policy evaluation is a fundamental challenge as driving quality is highly context-dependent.
arXiv:2606. 31106v1 Announce Type: cross Abstract: Large-scale datasets and fast simulators have enabled improvements in driving policies that appear safe and robust, yet strong performance in nominal scenarios can still mask flawed reasoning and unsafe heuristics.
OPTED is a method for on‑policy fine‑tuning of end‑to‑end driving models that separates reinforcement learning from the policy update. A privileged teacher trained with RL on vectorized inputs (HD‑maps and bounding boxes) supervises the pre‑trained student during closed‑loop post‑training. Applied to the camera‑based models TransFuser and VaVAM in AlpaSim, OPTED boosts driving scores by 1.6× and 9.5×, respectively, while requiring roughly three orders of magnitude fewer simulator interactions than direct RL post‑training.
arXiv:2609.22762v1 Announce Type: new Abstract: Generative world-action models (WAMs) jointly generate future video and vehicle actions, while their action branches remain primarily optimized by expe...
arXiv:2603. 28963v2 Announce Type: replace-cross Abstract: Simulation with realistic traffic agents is essential for validating autonomous driving systems.
REARL is a closed‑loop simulation enhancement framework that combines real traffic data with large language models (LLMs) to improve autonomous driving simulations. It clusters real traffic, uses cluster centers as representative scenarios for the LLM, and employs a sliding‑window detector to monitor vehicle speed and spacing discrepancies. When thresholds are exceeded, the LLM adjusts vehicle decision‑making or selects matching real vehicle actions, resulting in lower Hellinger distance and MAPE compared to baselines in a HighD highway setting.
arXiv:2607. 13028v1 Announce Type: cross Abstract: Training robust autonomous driving agents requires a simulator that is fast enough for reinforcement learning at scale, realistic enough to ground behavior in real-world map structure, and diverse enough to cover the safety-critical long tail that logged data rarely contains.
arXiv:2609.06433v1 Announce Type: cross Abstract: The generation of safety-critical traffic scenarios is essential for training and evaluating autonomous vehicles. Prior approaches typically perturb...
arXiv:2606. 03159v1 Announce Type: cross Abstract: As autonomous vehicle capabilities advance, the safe evaluation of driving policies in long-tail scenarios remains a critical bottleneck.