DynVLA: Learning World Dynamics for Action Reasoning in Autonomous Driving

• URL: http://arxiv.org/abs/2603.11041v2
• Date: Fri, 13 Mar 2026 01:57:24 GMT
• Title: DynVLA: Learning World Dynamics for Action Reasoning in Autonomous Driving
• Abstract summary: DynVLA is a driving VLA model that introduces a new CoT paradigm termed Dynamics CoT.<n>To obtain compact dynamics representations, DynVLA introduces a Dynamics Tokenizer that compresses future evolution into a small set of dynamics tokens.<n>We train DynVLA to generate dynamics tokens before actions through SFT and RFT, improving decision quality while maintaining latency-efficient inference.
• Score: 47.8729375581201
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We propose DynVLA, a driving VLA model that introduces a new CoT paradigm termed Dynamics CoT. DynVLA forecasts compact world dynamics before action generation, enabling more informed and physically grounded decision-making. To obtain compact dynamics representations, DynVLA introduces a Dynamics Tokenizer that compresses future evolution into a small set of dynamics tokens. Considering the rich environment dynamics in interaction-intensive driving scenarios, DynVLA decouples ego-centric and environment-centric dynamics, yielding more accurate world dynamics modeling. We then train DynVLA to generate dynamics tokens before actions through SFT and RFT, improving decision quality while maintaining latency-efficient inference. Compared to Textual CoT, which lacks fine-grained spatiotemporal understanding, and Visual CoT, which introduces substantial redundancy due to dense image prediction, Dynamics CoT captures the evolution of the world in a compact, interpretable, and efficient form. Extensive experiments on NAVSIM, Bench2Drive, and a large-scale in-house dataset demonstrate that DynVLA consistently outperforms Textual CoT and Visual CoT methods, validating the effectiveness and practical value of Dynamics CoT. Project Page: https://yaoyao-jpg.github.io/dynvla.

Related papers

• WOLF-VLA: Whole-Body Humanoid Optimal Locomotion Framework for Vision-Language-Action Learning [4.427336047705734]
  Vision-Language-Action (VLA) models have recently demonstrated strong generalization in robotic manipulation.<n>This work introduces a unified framework WOLF-VLA that integrates whole-body optimal-control (OC) motion synthesis with large-scale multi-modal dataset.<n>We construct a comprehensive dataset of dynamically feasible humanoid trajectories across six locomotion-related task families.<n>We train a VLA model using the collected joint trajectories, ego-centric visual observations and natural language instruction, yielding a policy that exhibits strong reasoning and robustness to initial-condition variability.
  arXiv  Detail & Related papers  (2026-06-24T08:59:59Z)
• Controlled Dynamics Attractor Transformer [58.15454173718687]
  Transformer architectures have advanced representation learning and inference in deep models through self-attention mechanisms.<n>In parallel,associative memory (AM) frameworks map representations onto energy landscapes, offering interpretable retrieval mechanisms.<n>We propose Controlled Dynamics Attractor Transformer (CCAT), which couples a mixture von Mises-Fisher (Mo-vMF) attention energy with a Hopfield refinement energy.<n>CCAT achieves state-of-the-art performance across multiple benchmarks in graph anomaly detection and graph classification.
  arXiv  Detail & Related papers  (2026-06-13T09:04:22Z)
• Physics-informed Deep Mixture-of-Koopmans Vehicle Dynamics Model with Dual-branch Encoder for Distributed Electric-drive Trucks [19.506735422845182]
  We propose a fully data-driven dynamics modeling method tailored for complex distributed electric-drive trucks (DETs)<n>We first propose a novel dual-branch encoder which encodes dynamic states and provides a powerful basis for the proposed Koopman-based methods entitled KODE.<n>A physics-informed supervision mechanism is incorporated into the training process to facilitate effective learning of both the encoder and the Koopman operator.
  arXiv  Detail & Related papers  (2026-03-18T06:47:40Z)
• Towards Generalizable Robotic Manipulation in Dynamic Environments [48.77270850350943]
  Vision-Language-Action (VLA) models excel in static manipulation but struggle in dynamic environments with moving trajectories.<n>We introduce DOMINO, a large-scale dataset and benchmark for general dynamic manipulation.<n>We also propose PUMA, a dynamics-aware VLA architecture.
  arXiv  Detail & Related papers  (2026-03-16T17:59:57Z)
• Chain of World: World Model Thinking in Latent Motion [24.24061036481793]
  Vision-Language-Action (VLA) models often overlook the predictive and temporal-causal structure underlying visual dynamics.<n>We introduce CoWVLA, a new "Chain of World" paradigm that unifies world-model temporal reasoning with a disentangled latent motion representation.<n>CoWVLA outperforms existing world-model and latent-action approaches and achieves moderate computational efficiency.
  arXiv  Detail & Related papers  (2026-03-03T17:52:06Z)
• DDP-WM: Disentangled Dynamics Prediction for Efficient World Models [79.53092337527382]
  We introduce DDP-WM, a novel world model centered on the principle of Disentangled Dynamics Prediction.<n>DDP-WM realizes this decomposition through an architecture that integrates efficient historical processing with dynamic localization.<n>Experiments demonstrate that DDP-WM achieves significant efficiency and performance across diverse tasks.
  arXiv  Detail & Related papers  (2026-02-02T08:04:25Z)
• Prismatic World Model: Learning Compositional Dynamics for Planning in Hybrid Systems [38.4555621948915]
  Prismatic World Model (PRISM-WM) is designed to decompose complex hybrid dynamics into composable primitives.<n>PRISM-WM significantly reduces rollout drift by accurately modeling sharp mode transitions in system dynamics.
  arXiv  Detail & Related papers  (2025-12-09T09:40:34Z)
• Vehicle Dynamics Embedded World Models for Autonomous Driving [13.012978896886134]
  We propose the Vehicle Dynamics embedded Dreamer (VDD) method, which decouples the modeling of ego-vehicle dynamics from environmental transition dynamics.<n>The proposed model significantly improves both driving performance and robustness to variations in vehicle dynamics, outperforming existing approaches.
  arXiv  Detail & Related papers  (2025-12-02T04:57:52Z)
• ImagiDrive: A Unified Imagination-and-Planning Framework for Autonomous Driving [64.12414815634847]
  Vision-Language Models (VLMs) and Driving World Models (DWMs) have independently emerged as powerful recipes addressing different aspects of this challenge.<n>We propose ImagiDrive, a novel end-to-end autonomous driving framework that integrates a VLM-based driving agent with a DWM-based scene imaginer.
  arXiv  Detail & Related papers  (2025-08-15T12:06:55Z)
• Towards Human-Like Trajectory Prediction for Autonomous Driving: A Behavior-Centric Approach [22.81464823797471]
  HiT (Human-like Trajectory Prediction) is a novel model designed to enhance trajectory prediction by incorporating behavior-aware modules and dynamic centrality measures.<n>To evaluate HiT's performance, we conducted extensive experiments using diverse and challenging real-world datasets.
  arXiv  Detail & Related papers  (2025-05-27T05:04:01Z)
• A PID-Controlled Tensor Wheel Decomposition Model for Dynamic Link Prediction [3.525733859925913]
  This study introduces a PID-controlled tensor wheel decomposition (PTWD) model, which mainly adopts the following two ideas.<n>The proposed PTWD model has more accurate link prediction capabilities compared to other models.
  arXiv  Detail & Related papers  (2025-05-20T11:14:30Z)
• Generative Pre-trained Autoregressive Diffusion Transformer [74.25668109048418]
  GPDiT is a Generative Pre-trained Autoregressive Diffusion Transformer.<n>It unifies the strengths of diffusion and autoregressive modeling for long-range video synthesis.<n>It autoregressively predicts future latent frames using a diffusion loss, enabling natural modeling of motion dynamics.
  arXiv  Detail & Related papers  (2025-05-12T08:32:39Z)
• Context-aware Dynamics Model for Generalization in Model-Based Reinforcement Learning [124.9856253431878]
  We decompose the task of learning a global dynamics model into two stages: (a) learning a context latent vector that captures the local dynamics, then (b) predicting the next state conditioned on it. In order to encode dynamics-specific information into the context latent vector, we introduce a novel loss function that encourages the context latent vector to be useful for predicting both forward and backward dynamics. The proposed method achieves superior generalization ability across various simulated robotics and control tasks, compared to existing RL schemes.
  arXiv  Detail & Related papers  (2020-05-14T08:10:54Z)

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.