On-Policy Distillation of Language Models for Autonomous Vehicle Motion Planning

• URL: http://arxiv.org/abs/2604.07944v1
• Date: Thu, 09 Apr 2026 08:06:19 GMT
• Title: On-Policy Distillation of Language Models for Autonomous Vehicle Motion Planning
• Abstract summary: Large language models (LLMs) have recently demonstrated strong potential for autonomous vehicle motion planning.<n>We study how to effectively transfer motion planning knowledge from a large teacher LLM to a smaller, more deployable student model.
• Score: 3.2748787252933442
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) have recently demonstrated strong potential for autonomous vehicle motion planning by reformulating trajectory prediction as a language generation problem. However, deploying capable LLMs in resource-constrained onboard systems remains a fundamental challenge. In this paper, we study how to effectively transfer motion planning knowledge from a large teacher LLM to a smaller, more deployable student model. We build on the GPT-Driver framework, which represents driving scenes as language prompts and generates waypoint trajectories with chain-of-thought reasoning, and investigate two student training paradigms: (i) on-policy generalized knowledge distillation (GKD), which trains the student on its own self-generated outputs using dense token-level feedback from the teacher, and (ii) a dense-feedback reinforcement learning (RL) baseline that uses the teacher's log-probabilities as per-token reward signals in a policy gradient framework. Experiments on the nuScenes benchmark show that GKD substantially outperforms the RL baseline and closely approaches teacher-level performance despite a 5$\times$ reduction in model size. These results highlight the practical value of on-policy distillation as a principled and effective approach to deploying LLM-based planners in autonomous driving systems.

Related papers

• Orion-Lite: Distilling LLM Reasoning into Efficient Vision-Only Driving Models [16.98212985592303]
  Distilling Large Language Models (LLMs) into a compact driving model offers a compelling solution.<n>We demonstrate that an efficient vision-only student model textbfOrion-Lite can even surpass the performance of its massive VLA teacher, ORION.
  arXiv  Detail & Related papers  (2026-04-09T13:51:55Z)
• PILOT: Planning via Internalized Latent Optimization Trajectories for Large Language Models [51.43746425777865]
  Large Language Models (LLMs) often lack the capacity to formulate global strategies, leading to error propagation in long-horizon tasks.<n>We propose PILOT, a framework designed to internalize the strategic oversight of large models into intrinsic Latent Guidance.
  arXiv  Detail & Related papers  (2026-01-07T12:38:56Z)
• Less is More: Lean yet Powerful Vision-Language Model for Autonomous Driving [7.921556303360947]
  We introduce Max-V1, a novel framework for one-stage end-to-end autonomous driving.<n>Our framework presents a single-pass generation paradigm that aligns with the inherent sequentiality of driving.<n> Empirically, our method achieves the state-of-the-art performance on the nuScenes dataset.
  arXiv  Detail & Related papers  (2025-09-29T05:14:18Z)
• Discrete Diffusion for Reflective Vision-Language-Action Models in Autonomous Driving [55.13109926181247]
  We introduce ReflectDrive, a learning-based framework that integrates a reflection mechanism for safe trajectory generation via discrete diffusion.<n>Central to our approach is a safety-aware reflection mechanism that performs iterative self-correction without gradient.<n>Our method begins with goal-conditioned trajectory generation to model multi-modal driving behaviors.
  arXiv  Detail & Related papers  (2025-09-24T13:35:15Z)
• ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving [49.07731497951963]
  ReCogDrive is a novel Reinforced Cognitive framework for end-to-end autonomous driving.<n>We introduce a hierarchical data pipeline that mimics the sequential cognitive process of human drivers.<n>We then address the language-action mismatch by injecting the VLM's learned driving priors into a diffusion planner.
  arXiv  Detail & Related papers  (2025-06-09T03:14:04Z)
• TeLL-Drive: Enhancing Autonomous Driving with Teacher LLM-Guided Deep Reinforcement Learning [61.33599727106222]
  TeLL-Drive is a hybrid framework that integrates a Teacher LLM to guide an attention-based Student DRL policy.<n>A self-attention mechanism then fuses these strategies with the DRL agent's exploration, accelerating policy convergence and boosting robustness.
  arXiv  Detail & Related papers  (2025-02-03T14:22:03Z)
• From Words to Actions: Unveiling the Theoretical Underpinnings of LLM-Driven Autonomous Systems [59.40480894948944]
  Large language model (LLM) empowered agents are able to solve decision-making problems in the physical world. Under this model, the LLM Planner navigates a partially observable Markov decision process (POMDP) by iteratively generating language-based subgoals via prompting. We prove that the pretrained LLM Planner effectively performs Bayesian aggregated imitation learning (BAIL) through in-context learning.
  arXiv  Detail & Related papers  (2024-05-30T09:42:54Z)
• Prompting Multi-Modal Tokens to Enhance End-to-End Autonomous Driving Imitation Learning with LLMs [10.812418229495506]
  We propose a hybrid End-to-End learning framework for autonomous driving by combining basic driving imitation learning with Large Language Models. The proposed methodology can attain driving scores of 49.21%, coupled with an impressive route completion rate of 91.34% in the offline evaluation conducted via CARLA.
  arXiv  Detail & Related papers  (2024-04-07T08:31:12Z)
• GPT-Driver: Learning to Drive with GPT [47.14350537515685]
  We present a simple yet effective approach that can transform the OpenAI GPT-3.5 model into a reliable motion planner for autonomous vehicles. We capitalize on the strong reasoning capabilities and generalization potential inherent to Large Language Models (LLMs) We evaluate our approach on the large-scale nuScenes dataset, and extensive experiments substantiate the effectiveness, generalization ability, and interpretability of our GPT-based motion planner.
  arXiv  Detail & Related papers  (2023-10-02T17:59:57Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.