Digest

Reset

1167 papers

2026-10-20 · arXiv:2610.25600

Vector-Quantized World Model: Latent Feedback Motion Planning for Robot Manipulation

Xuezhou Xu, Chung-Pang Wang, Zhenzhen Li et al.

VQ-WM is a latent feedback motion planning framework for long-horizon robot manipulation. It combines discrete search over vector-quantized latent states with a continuous visual-feedback controller that sequentially reaches planned targets. State vector quantization is based on first-passage time (controller calls to enter a neighborhood), aligning the planning abstraction with controller reachability. The authors report 100% success on physical Tower of Hanoi with only short-play data.

RoboticsAI summary
2026-10-20 · arXiv:2610.25666

Frechet Forcing: Training Long-Horizon Streaming World Models from Short Clips

Yian Zhao, Yu Yan, Hongcan Guo et al.

Fréchet Forcing fine-tunes a streaming action-conditioned video world model using only short real clips. It trains the deployed multi-step sampler with three Fréchet-distance constraints: flow-matching after a self-generated history that ends on real data, context fidelity via clip-centred latent statistics, and reality matching via Fréchet distance in a frozen perceptual space. For MiniWorld-1B, fine-tuning on 381-frame clips yields stable rollouts beyond 4,000 frames while preserving action following.

Video World ModelsCode availableAI summary
2026-10-20 · arXiv:2610.25690

ScaleNav: Scalable Internet Video Pretraining for Language-Conditioned Local Navigation

Yihui Yang, Yuanhong Yu, Jiawei Wang et al.

ScaleNav targets language-conditioned local navigation to a visible text-specified target. It learns goal-directed visual dynamics from action-free Internet egocentric videos via instruction-conditioned video pretraining, aligns action prediction using action-labeled simulation, then jointly fine-tunes. The unified world-action model predicts future observations and executable action chunks, and is evaluated in simulation and on a real quadruped robot.

DrivingAI summary
2026-10-20 · arXiv:2610.25734

Think Densely, Act Sparsely: Latent Expert Cognitive Chains for Vision-Language-Action Autonomous Driving

Jie Wang, Guang Li, Zhijian Huang et al.

The authors propose LECDrive under a “Think Densely, Act Sparsely” paradigm for vision-language-action autonomous driving. LECDrive injects compact task-specific latent tokens into a VLM (Qwen2.5-VL-7B) as a Latent Expert Cognitive Chain with three stages: DINOv3 (visual), SAM3 (semantic), and DepthAnything3 (geometry). Dense world modeling is trained via dedicated dense perception heads, while actions remain sparse.

DrivingAI summary
2026-10-20 · arXiv:2610.25763

TrajWAM: An Efficient Trajectory-Guided World Action Model for Dynamic Robotic Manipulation

Haohan Geng, Sha Zhang, Weishuo Sun et al.

TrajWAM is a trajectory-guided world action model for dynamic robotic manipulation. It predicts future target-object masks from a history of object masks, uses a lightweight Trajectory ControlNet to steer a pretrained video model toward trajectory-consistent futures, and conditions an action expert on the resulting future representation. A decoupled one-step distillation enables efficient one-step video-action inference for closed-loop control.

DrivingAI summary
2026-10-20 · arXiv:2610.25764

ActTime: Action-Conditioned World Modeling for Time Series

Xinxing Zhou, Xiaojie Yuan, Yanlong Wen et al.

ActTime is an action-conditioned world model for time series that learns action-conditioned dynamics from factual trajectories only. It updates variable-specific latent states via an action-conditioned transition plus observation correction, then rolls the same transition forward under candidate future actions to generate alternative-action trajectories. In simulator-grounded evaluations across four domains, it ranks first in 12 of 16 task settings and shows that factual accuracy, intervention fidelity, and decision quality are related but not interchangeable.

Latent DynamicsAI summary
2026-10-20 · arXiv:2610.25781

See the Future, Classify the Present: Stochastic Latent World Modeling for Hyperspectral Images

Muhammad Ahmad, Gemine Vivone, Ronny H\"ansch et al.

Stochastic Latent World Modeling (LWM) for hyperspectral image classification replaces static spatial-spectral embeddings with a trajectory-aware latent representation. It tokenizes spectral (per band) and spatial (per pixel) information using transformer branches, builds an environment state, then uses a stochastic latent transition model (Gaussian parameterization + recurrent refinement) to predict multi-step latent rollouts. A rollout planner adaptively aggregates the trajectory for classification.

Latent DynamicsAI summary
2026-10-20 · arXiv:2610.25861

ActionRoPE: action as coordinates for spatially consistent control in planar-camera game world models

Zhiyang Deng, Yuzhe Wu, Zijun Lin et al.

ActionRoPE represents action effects as world-coordinate rotary positional encoding in planar-camera game world models, so tokens at the same world location keep the same positional address across time. It enables spatially consistent control. The authors add CNS (Coordinate-Conditioned Noise Scheduling) to assign token-wise denoising schedules based on geometrically known vs newly revealed regions, stabilizing revisits and reducing drift.

DrivingAI summary
2026-10-20 · arXiv:2610.25874

Vision-Language-Delta-Action: Learning Compact Deltas as Actionable Transition Interfaces for Robotic Manipulation

Lingsen Zhang, Wei Li, Jie He et al.

VLDA (Vision-Language-Delta-Action) reframes robotic manipulation from observation-to-action prediction to an understanding–transition–execution loop. It introduces a compact actionable Delta latent capturing task-specified physical state changes via a unified objective (LT-Filter + GT-Anchor), then uses Delta-Driven Action Refinement (DDAR) to seed action denoising, aiming to avoid wasting capacity on full future-state prediction.

RoboticsCode availableAI summary
2026-10-20 · arXiv:2610.25893

Motor compound fault diagnosis using an attention-based cross-talk architecture

Wonjun Yi, Wonho Jung, Junho Kim et al.

The authors propose RNDR-TFCA, an attention-based cross-talk architecture for multi-output compound fault diagnosis in motors. It keeps task-specific feature extractors while selectively exchanging features across tasks using time-frequency attention (intra-task) and channel attention via an SE block (inter-task). On a rotor-bearing system, RNDR-TFCA achieves a macro F1 of 0.898, outperforming RNDR and other baselines.

DrivingAI summary
2026-10-20 · arXiv:2610.25959

EmbodiedWorld-200K: Toward Embodied World Agents via Embodied-Planning Dataset and Interactive World Models

X. Feng, J. Tang, Z. Lin et al.

The authors release EmbodiedWorld-200K (over 200K samples from 150 3D open-world games) for open-world embodied navigation-planning, with multi-level annotations and unified evaluation. They pretrain a VLA EWA offline on the dataset, then use a video world model as a low-cost environment proxy for interactive RL post-training with VQA-based reward scoring and DPO. They report improved performance on in-distribution and OOD evaluation splits.

2026-10-20 · arXiv:2610.25961

PaceVLA: Adaptive Temporal Action Quantization in Vision-Language-Action Models

Hojin Jeon, Jimin Lee, Taekwan Kim et al.

PaceVLA is an adaptive vision-language-action framework for temporally efficient robot execution. It uses a shared flow-matching action expert with mode-specific decoders for multiple temporal granularities and prediction horizons, plus a quantizability-guided mode selection (quantization judge + context-conditioned router) to choose fine vs quantized actions per inference step.

DrivingAI summary
2026-10-20 · arXiv:2610.26179

Learning Goal-Conditioned Subgoals through Frozen Hierarchical World Models

Xiang Kaibo, Liu Qingfeng

The authors replace expensive high-level candidate optimization in Hierarchical Planning with Latent World Models (HWM) with a lightweight goal-conditioned head trained on offline trajectories. The head is trained through a frozen high-level predictor using teacher-forced, rollout, and endpoint losses from hindsight goals, producing the first latent waypoint (subgoal) for a retained low-level MPC planner. It removes high-level candidate optimization at deployment.

DrivingAI summary
2026-10-20 · arXiv:2610.26390

Cartender: Verifiable Language-Conditioned Driving by Learning the Drives Not Taken

Zhiwen Chen, Benjin Zhu, Yan Xie et al.

Cartender learns verifiable language-conditioned driving from logged data where each scene is recorded once. It uses Counterfactual Branching to construct matched/mismatched instruction-action pairs from real recordings (no simulator), then gates language conditioning with a learned scene-language compatibility score and ranks candidates with a reward-trained language residual. Evaluation on WOD-E2E uses behavior under instruction swapping.

DrivingAI summary
2026-10-20 · arXiv:2610.26564

PrismWAM: Decoupling State Representations in World Action Models

Tianyu Xin, Wanqi Zhong, Bowen Hu et al.

PrismWAM proposes decoupling semantic context from physical state in World Action Models. It factorizes embodied control into a Semantic Planning → Physical Rollout → Executable Action pipeline: from initial RGB it predicts a goal image (sparse), while a depth-based rollout predicts dense geometric evolution (dense) and generates continuous actions, enabling strong manipulation in simulation and the real world without embodied action pretraining.

DrivingAI summary
2026-10-20 · arXiv:2610.26711

PVBench: Quantifying Physical Plausibility in Video Generation with Object-Level Physical Measurements

Kaiwei Chen, Moran Li, Zhenye Gan et al.

PVBench is a benchmark for physical plausibility in video generation. It evaluates 46 dimensions across seven domains and three layers (semantic, perceptual, physical) using specialized vision models that extract object-level physical quantities (e.g., 6D pose, metric scale, surface normals, kinematics, conservation-law residuals), with MLLM scoring where quantities are hard to compute. The authors release code, prompts, and a calibration protocol.

Video World ModelsCode availableAI summary
2026-10-20 · arXiv:2610.26757

Simulation Priors for Rapid Adaptation in Model-Based RL for Autonomous Racing

Adrian Remonda, Jiajun Xi, Nicklas Hansen et al.

PIA (Prior-Informed Adaptation) is an offline-to-online model-based RL framework for autonomous racing that pretrains a world model on heterogeneous low- and high-fidelity simulation trajectories, then adapts with limited real-world interaction using uncertainty-regularized latent planning. The authors treat simulated trajectories as biased priors and correct residual mismatch during online adaptation, avoiding the need for an exact simulator. They evaluate in ACGym and on a full-size autonomous race car.

DrivingAI summary
2026-10-20 · arXiv:2610.26780

V-JEPA Hand: Simple, Accurate and Fast 3D Hand and Camera Trajectory Recovery from Egocentric Videos

Ruicheng Li, Qixiu Li, Yu Deng et al.

V-JEPA Hand is a single feed-forward video model that jointly reconstructs bilateral 3D hand motion and camera egomotion from egocentric clips in one forward pass. Using a self-supervised V-JEPA 2.1 encoder plus a query-based predictor, it outputs per-frame hand poses/presence and camera motion, expressing hand trajectories in a common world frame. The model accepts known focal lengths or predicts them and is trained end to end on mixed pseudo-labelled and high-precision 3D annotations.

JEPAAI summary
2026-10-20 · arXiv:2610.26811

State World Models: Execute the Rules, Generate the Pixels

Chan Yan Ki, Yanran Zhang, Wenzhao Zheng et al.

The authors propose Stateful Harness, which separates rule execution/state evolution from video rendering. It maintains evolving semantic/spatial state under fixed rules, renders each resolved event using a frozen video model, and uses the generated clip to update state for the next turn. They evaluate rule compliance with WorldState-Bench (50 counterfactual twin pairs with hidden state), where the full harness reaches 92% PairFollow.

2026-10-20 · arXiv:2610.26813

PICA-WAM: Visual Perturbation Invariance via Cross-Appearance Self-Alignment in World Action Models

Yude Zou, Yifei Yao, Guanbo Jiao et al.

PICA-WAM trains World Action Models to be invariant to task-irrelevant visual perturbations by aligning action-stream representations across appearances. An EMA teacher encodes the nominal observation at a lower noise level, and the online model encodes a perturbed observation; a projection head aligns the student’s shallower action features to the teacher’s deeper ones. The authors report robustness gains on RoboTwin 2.0, LIBERO-Plus, EBench, and a real robot, while preserving nominal performance.

DrivingCode availableAI summary