Digest

Reset

1167 papers

2026-10-28 · arXiv:2610.36791

RynnWorld-Latent: A Cross-Embodiment World Model Conditioned on Universal Latent Actions

Haoyu Zhao, Yinzhou Tang, Zixuan Wang et al.

RynnWorld-Latent is a cross-embodiment, text-free video world model conditioned on 3D-grounded continuous latent actions from RynnLAM. It jointly pretrains on 23 datasets (~553.5M frames) and is post-trained to target robots. The model supports (1) digital teleoperation to generate action-video data and (2) latent-action synthesis via first-frame editing for cross-embodiment rollouts, improving prediction and policy success in evaluated settings.

Latent DynamicsAI summary
2026-10-28 · arXiv:2610.36823

TAPESTRY: A Single Backbone for Video Generation and Robot Control

Jeremy A. Collins, Seungjae Lee, Rachanon Wachakorn et al.

TAPESTRY repurposes a pretrained text-conditioned video diffusion transformer as an action-generating policy by treating each robot action timestep as a single “patch” token in the video VAE latent space. Actions are projected with a lightweight linear layer + affine-free LayerNorm, then processed by the shared transformer. The model is trained with video, joint video+action, and action-only diffusion, and deployed with one DDIM step (no video generation).

RoboticsAI summary
2026-10-28 · arXiv:2610.36840

Trace-WAM: Learning to Act via 3D Motion Prediction for World Action Models

Xu Hu, Changjie Wu, Linzhe Shi et al.

Trace-WAM is a World Action Model that augments video-action learning with future 3D motion prediction as an auxiliary objective. It supervises temporal point correspondence via image-plane displacement and depth change, and uses a three-branch Tri-MoT architecture to jointly learn future video prediction, 3D motion prediction, and action generation. At inference, it follows the Fast-WAM policy path using only current observations.

DrivingAI summary
2026-10-28 · arXiv:2610.36877

GUI-JEPA: Candidate-Conditioned JEPA Latents as a Fast Action Policy for Mobile GUI Agents

Bingchao Wang, Yongcheng Jing, Shunyu Liu et al.

GUI-JEPA is a decoder-free mobile GUI agent policy built on a JEPA encoder pretrained on unlabeled GUI screenshots. It embeds the screenshot once, uses an action- and region-conditioned latent predictor to form one latent per candidate action, and scores all candidates in a single batched pass (no token decoding). On AndroidControl, it is 2.8–4.0× faster than a same-protocol 7B VLM baseline and reaches 86–92% of its in-distribution accuracy with 6% of its parameters.

DrivingCode availableAI summary
2026-10-28 · arXiv:2610.36925

Vision-Language Agents Can Learn in Their Own Latent Worlds

Haoxi Li, Sikai Bai, Jianfei Ma et al.

Dwell is a model-based RL framework that reframes a pretrained vision-language model (VLM) as a recurrent latent transition operator. It uses separate lightweight LoRA adapters for world modeling and policy while sharing the same latent state interface, enabling online learning from real interaction plus latent imagination. Across five interactive visual environments, the authors report the strongest overall performance among trainable methods while reducing imagination compute.

Latent DynamicsCode availableAI summary
2026-10-28 · arXiv:2610.37091

GAWM: A Compact Latent World-Action Model for Multi-Task Robot Manipulation

Yang Liu, Xiaoli Liu, Liuzhao Li et al.

GAWM is a compact, task-conditioned latent world-action model for multi-task robot manipulation. It uses a frozen DINOv3 encoder to compress each camera view into 64 tokens, forecasts future tokens at two horizons in a task-conditioned predictor, and decodes an action chunk from current and predicted tokens. With 77M params and no robot/video pretraining, GAWM-S outperforms billion-parameter WAMs on RoboTwin 2.0 and is comparable on LIBERO, with lower real-robot inference latency.

Latent DynamicsCode availableAI summary
2026-10-28 · arXiv:2610.37175

MLE-World: Predicting Machine Learning Experiment Outcomes with a Learned World Model

Brian Ko, Han Wang, Hejia Geng et al.

MLE-World is a learned world model fine-tuned from Qwen3.5-4B to predict a program’s terminal output (error traces, training logs, validation score) from its code and data description without executing it, plus a confidence score. It enables test-time scaling for MLE agents via five world-model-guided policies with a confidence gate for selective real execution, improving held-out matching and observed mean performance while reducing real executions.

Latent DynamicsAI summary
2026-10-28 · arXiv:2610.37196

How Much Can Best-of-N Help a World Action Model? Measuring Selectable Headroom and Realizing It in Closed Loop

Haojia Gao, Yue Cao, Yumin Liu et al.

The authors argue that standard best-of-N selectable-headroom estimates confound candidate quality with randomness of future continuation. They introduce a denoised counterfactual branching protocol with repeated continuations, held-out oracles, and common random numbers to measure the selectable share. Using four WAMs, they find selection headroom depends strongly on structured action proposals, and an imagination-based ranker with a fail gate improves closed-loop task success.

DrivingAI summary
2026-10-28 · arXiv:2610.37215

Retrieve to Act: Motion Primitive Graph Retrieval for Robust VLA Execution

Pham Tri Quang, Bao-Ngoc Dao, Minh Duc Nguyen et al.

Retrieve to Act (R2A) is a test-time scaling method for robust VLA execution under OOD visual shifts. It builds an offline motion primitive graph from successful demonstrations using only kinematic geometry, then uses a query-dependent GNN to retrieve structurally compatible future motion primitives. A learned dimension-wise gating module fuses retrieved priors with frozen VLA actions, improving robustness without finetuning.

DrivingAI summary
2026-10-28 · arXiv:2610.37219

Vision-Language-Action or World-Action? Just Action Models

Yonghao Yu, Lang Huang, Runyi Li et al.

The authors propose Just Action Model (JAM): train a pretrained visual backbone with an action expert using only an action loss plus a “foresight” loss that predicts frozen encodings of recorded future observations via a training-only head and anchor branch. No visual generation is used in training or inference. JAM outperforms matched FastWAM and ImageWAM on LIBERO/LIBERO-Plus and matches or slightly improves on RoboTwin; caching foresight targets reduces online training FLOPs per sample per update.

DrivingAI summary
2026-10-28 · arXiv:2610.37257

GameVerse: A Minute-Scale Gameplay Dataset for Long-Horizon Interactive World Modeling

Kang He, Wenshuo Peng, Chuanhao Li et al.

GameVerse is a minute-scale gameplay video dataset for long-horizon interactive world modeling. The authors report 69.7K clips (1,655 hours, 349.6M frames), including 1,358 hours of continuous two-minute sequences, with temporally stable frame-level camera trajectories and four-level event-aligned captioning (global static, local static, dynamic events, agent actions). They also report that training on GameVerse improves long-horizon temporal consistency and visual quality on a 30-second benchmark.

2026-10-28 · arXiv:2610.37381

PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representations

Hanqing Wang, Zhaotian Cai, Ruichuan An et al.

PDRVLA is a vision-language-action model that adds a lightweight Physical Dynamics Module to extract physical tokens from intermediate vision-language features. The physical module is trained to predict future RGB, depth, and optical-flow features, with auxiliary supervision from a frozen pretrained video world model. The learned physical tokens condition action generation without online future-video generation. Trained on publicly available data expanded via human-to-robot synthesis and annotations.

DrivingAI summary
2026-10-28 · arXiv:2610.37391

JEPA-Inspired Latent Field Prediction for Sparse-Pilot MIMO-OFDM Channel Reconstruction

Yuzhi Yang, Brahim Mefgouda, Hang Zou et al.

The authors propose a JEPA-inspired field world model (FWM) for sparse-pilot MIMO-OFDM channel reconstruction. CSI is tokenized into latent time-frequency tokens; the model predicts masked/future latent field states from historical multi-band context, then fuses the predicted latent prior with sparse noisy pilots to reconstruct full CSI. Ray-tracing experiments show improved NMSE over pilot-only and direct reconstruction baselines.

JEPAAI summary
2026-10-28 · arXiv:2610.37437

Representing the Unseen: Rethinking Visual Imagination for Spatial Reasoning

Zengqi Zhao, Zun Wang, Shoubin Yu et al.

The authors argue that spatial reasoning over imagined states should not default to RGB. They introduce ImagiRep, a simulator-grounded benchmark with oracle RGB and six alternative representations of the same unobserved state, and show that accurate imagined RGB does not guarantee accurate spatial reasoning and that different tasks benefit from different representations. They further propose GEARS, a test-time adaptive representation selection strategy that outperforms existing methods.

Latent DynamicsAI summary
2026-10-28 · arXiv:2610.37518

World Dreamer v5-lite: A Multi-Robot World Action Model with One Dreamer

Guoning Wu (World Agents), Zijian Cai (World Agents), Yuhang Zhang (World Agents) et al.

World Dreamer v5-lite is a multi-robot world-action model built around a shared world state (ϕ) in workspace coordinates. Calibrated features from a pretrained video encoder populate ϕ, while a vision-language model provides scene context and one query per robot. Queries exchange pose-aware information, but action streams remain separate. Masked-view cooperative training withholds robot views while keeping action targets.

DrivingAI summary
2026-10-28 · arXiv:2610.37567

Gym-Anything-RT: Scaling Real-Time Environments for Vision-Language Agents

Pranjal Aggarwal, Sean Welleck

Gym-Anything-RT builds 100 visual minigames (and non-real-time counterparts) to test vision-language agents under real-time interaction. The same minigame can be run live, slowed, or paused to separate real-time delays from underlying task difficulty. The authors report large capability gaps: GPT-6 Astra (xHigh) is 9.0% pass in real time vs 75.0% when paused, and they publicly release the benchmark and generated minigames.

DrivingCode availableAI summary
2026-10-28 · arXiv:2610.37588

Understanding the limits of latent world models

Rodrigo Gonzalez Laiz, Steffen Schneider

The authors study when latent world models learn “faithful latent dynamics,” defined as an equivariance relation between the environment’s transition map and learned latent dynamics, plus state sufficiency. They introduce latent rollout consistency (LRC) and post-hoc state decoding. Across DreamerV3, TD-MPC2, and LeWM on Pendulum, CartPole, and Reacher, the authors report none solves tasks by actually faithfully modeling environment dynamics.

Latent DynamicsAI summary
2026-10-28 · arXiv:2610.37666

ThinkWorld: World Imagination Reinforces Chain-of-Dynamics Reasoning

Chaehyun Kim, Chi-Pin Huang, Jisoo Kim et al.

ThinkWorld is a VLA framework that grounds action reasoning in world dynamics using Chain-of-Dynamics (CoD): a sequence of discrete motion tokens. It initializes CoD with action-labeled demonstrations, then predicts future representations in a frozen DINO feature space and uses consequence-grounded reinforcement learning (rewarding imagined vs observed futures) to learn from videos without motion annotations, enabling human-to-robot transfer and generalization.

DrivingAI summary
2026-10-27 · arXiv:2610.33862

Securing Edge & Physical AI: Post-Quantum IoT Blockchain for Mission-Critical Cyber-Physical Agents

K\"ur\c{s}at \c{C}akal, Suat \"Ozdemir

The authors propose “rotational secrecy” for edge/physical AI systems under quantum threats, aiming to limit temporal key exposure via authenticated transitions between cryptographic epochs. They introduce PARS (Private Authentication for Rotational Secrecy) for policy-triggered key establishment/rotation/reconciliation, and an optional permissioned PARS-Enabled Edge Security Ledger (PARS-ESL) that records authenticated epoch/state (not high-rate transactions) for lightweight clients.

DrivingAI summary
2026-10-27 · arXiv:2610.33966

InteractiveWorld: A Verifiable Human-Centric Environment for Building Better Interactive MLLM Assistants

Qinsi Wang, Wei Wen, Chong Zhou et al.

InteractiveWorld is a verifiable, human-centric simulator for closed-loop interaction between an assistant and a human who executes the assistant’s natural-language guidance. It introduces InteractiveBench, a dynamic benchmark with 120 multi-turn egocentric tasks across six categories, scored by verifiable state-based rewards. Experiments with 12 frontier MLLMs show large gaps: Claude Fable 5.1 reaches 41.7% average success, with only 20.0% on Guidance and 25.0% on Planning.

DrivingCode availableAI summary