# Aligning Multi-Trajectory Supervision with Policy Optimization for VLA Driving

arXiv: https://arxiv.org/abs/2608.30122 · HTML: https://mlsanity.com/paper/2608.30122/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Chain-of-thought reasoning for driving, one of 17 directions in LLM reasoning.
- Builds on 21 of its 26 references that are in the map.
- First posted on arXiv 31 Aug 2026.

## Placed in

- LLM reasoning / [Chain-of-thought reasoning for driving](https://mlsanity.com/lr/chain-of-thought-reasoning-for-driving/)
- 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/)
- Offline RL and imitation learning from data / [Imitation and offline RL for driving](https://mlsanity.com/offlinerl/imitation-and-offline-rl-for-driving/)

## Builds on (most cited first, 21 of 21)

- [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025)
- [Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction](https://mlsanity.com/paper/1906.00949/) (2019)
- [AWAC: Accelerating Online Reinforcement Learning with Offline Datasets](https://mlsanity.com/paper/2006.09359/) (2020)
- [TransFuser: Imitation with Transformer-Based Sensor Fusion for Autonomous Driving](https://mlsanity.com/paper/2205.15997/) (2022)
- [AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning](https://mlsanity.com/paper/2506.13757/) (2025)
- [Hydra-MDP: End-to-end Multimodal Planning with Multi-target Hydra-Distillation](https://mlsanity.com/paper/2406.06978/) (2024)
- [EMMA: End-to-End Multimodal Model for Autonomous Driving](https://mlsanity.com/paper/2410.23262/) (2024)
- [OpenDriveVLA: Towards End-to-end Autonomous Driving with Large Vision Language Action Model](https://mlsanity.com/paper/2503.23463/) (2025)
- [Enhancing End-to-End Autonomous Driving with Latent World Model](https://mlsanity.com/paper/2406.08481/) (2024)
- [FutureSightDrive: Thinking Visually with Spatio-Temporal CoT for Autonomous Driving](https://mlsanity.com/paper/2505.17685/) (2025)
- [OmniDrive: A Holistic Vision-Language Dataset for Autonomous Driving with Counterfactual Reasoning](https://mlsanity.com/paper/2504.04348/) (2025)
- [DriveVLA-W0: World Models Amplify Data Scaling Law in Autonomous Driving](https://mlsanity.com/paper/2510.12796/) (2025)
- [AlphaDrive: Unleashing the Power of VLMs in Autonomous Driving via Reinforcement Learning and Reasoning](https://mlsanity.com/paper/2503.07608/) (2025)
- [Hydra-MDP++: Advancing End-to-End Driving via Expert-Guided Hydra-Distillation](https://mlsanity.com/paper/2503.12820/) (2025)
- [AdaThinkDrive: Adaptive Thinking via Reinforcement Learning for Autonomous Driving](https://mlsanity.com/paper/2509.13769/) (2025)
- [From Forecasting to Planning: Policy World Model for Collaborative State-Action Prediction](https://mlsanity.com/paper/2510.19654/) (2025)
- [Drive-JEPA: Video JEPA Meets Multimodal Trajectory Distillation for End-to-End Driving](https://mlsanity.com/paper/2601.22032/) (2026)
- [Plan-R1: Safe and Feasible Trajectory Planning as Language Modeling](https://mlsanity.com/paper/2505.17659/) (2025)
- [Devil is in Narrow Policy: Unleashing Exploration in Driving VLA Models](https://mlsanity.com/paper/2603.06049/) (2026)
- [HAD: Combining Hierarchical Diffusion with Metric-Decoupled RL for End-to-End Driving](https://mlsanity.com/paper/2604.03581/) (2026)
- [HMAD: Advancing E2E Driving with Anchored Offset Proposals and Simulation-Supervised Multi-target Scoring](https://mlsanity.com/paper/2505.23129/) (2025)

## Abstract (arXiv)

Vision-language-action (VLA) driving methods increasingly combine multi-trajectory imitation learning with group-relative policy optimization (GRPO), making trajectory selection critical to final performance. However, some high-scoring trajectories that improve imitation can degrade subsequent GRPO by inducing advantage estimates misaligned with the current policy's feasible behavior distribution, driving updates away from safe and compliant behaviors. To address this, we propose a novel framework that aligns multi-trajectory supervision with policy optimization. To address the policy gradient bias induced by infeasible noisy trajectories outside the feasible region, augmented trajectories are constrained to a neighboring manifold of the ground-truth feasible region, and a Pareto-optimality criterion is adopted in place of the conventional aggregate score, retaining only non-dominated candidates and thereby filtering out conflicting samples at the source. To ensure that expanded trajectory supervision is effectively absorbed during policy optimization, we introduce two complementary mechanisms: feasibility-first advantage assignment and dynamic distillation. The former adapts Pareto credit to the feasibility composition of each rollout group and guides fully infeasible groups toward safe references. The latter updates teacher trajectories across refinement rounds to continually transfer useful supervision. Together, they progressively translate the benefits of expanded supervision into policy improvement. On NAVSIM v1 and v2, our method achieves 91.4 PDMS and 89.1 EPDMS, respectively, under single-trajectory inference, and recovers 440 of 658 initially failed scenes, 11.1% higher than the original GRPO baseline.