世界模型 前沿增量 · 2026-10¶
生成：
py -3.9 tools/watch.py --volume wm --write｜窗口 2026-06-11 ~ 2026-10-09（120 天）｜词表 24 条｜请求失败 0 条｜新增 346 条
词表按 docs/02-世界模型专题/07–09 三篇 分节。
读法（纪律 F 的镜像规则）：TOTAL 是题摘层面的命中数，不是相关论文数。一条宽查询命中几百条不含任何信息，所以下表只列仓库尚未收录的条目。NULL 表示请求失败（429/超时），不是 0 命中；0 表示确实没有，但第一反应应当是怀疑查询措辞，换个提法复核后再下结论。
arXiv 的 all: / abs: 只覆盖题名/摘要/作者/注释，不覆盖全文。
去重范围是整仓：docs/、references/、mindmaps/、paper/ 下的 .md，加 README.md / CONTRIBUTING.md / code/README.md。paper/ 下的 _work*/ 取数缓存（json/xml/html/pdf）不算收录，故不计入——那里放的是原始查询结果，不是读过的文献。所以某条 ID 不在本表，只能推出整仓没写过，推不出"这个专题没有"。
（2026-10-08 修：paper/ 原先漏扫，加上 ID 正则不吃 v1 版本号后缀，两条合起来让 27 条早已在 paper/ 笔记里读过的论文被标成"新增"。三卷的台账互为已收录，故 <卷>-watch-*.md 一律排除。）
07 · 世界模型评测与诊断¶
abs:"world model" AND abs:"evaluation" — 命中 471，其中新增 21（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.09763v1 |
Beyond Policy Support: Interaction Constrained Offline Reinforcement Learning for Autonomous Driving |
| 2026-10-07 | 2610.09546v1 |
PCDT: A Predictive Cognitive Digital Twin Framework for Intelligent and Autonomous 6G Network Ecosystems |
| 2026-10-07 | 2610.09438v1 |
Controllable Crowd Generation through World-Model Planning |
| 2026-10-07 | 2610.09307v1 |
vLLM-Omni Technical Report: A Unified Serving Runtime for Omni-Modality Generation |
| 2026-10-07 | 2610.09305v1 |
Kuration SDK: Addressing the Virtual2Real Gap via Data Curation |
| 2026-10-06 | 2610.09048v1 |
Towards Financial World Modeling |
| 2026-10-06 | 2610.08780v1 |
DepthWorld: 3D World Model for Robot Manipulation |
| 2026-10-06 | 2610.08777v1 |
CtrlCache: Accelerating Interactive Video World Models with Control-Aware Caching |
| 2026-10-06 | 2610.08640v1 |
RIWANav: Recursive World-Action Models with Self-Improvement for Urban Navigation |
| 2026-10-06 | 2610.08469v1 |
A Belief-State World Model for Catheter Navigation under Sparse Fluoroscopy: A Planar Proof of Concept |
| 2026-10-06 | 2610.08464v1 |
Federated Bayesian Surveillance of Mechanical Thrombectomy Adverse Events: A Population Risk Layer for Surgica |
| 2026-10-06 | 2610.08267v1 |
WM4ISAC: World Model for Proactive ISAC Under Dynamic Blockage |
| 2026-10-06 | 2610.08033v1 |
Learning in Dreams, Winning in Reality: A Continuous Dyna Loop for a Ten-Hero MOBA |
| 2026-10-06 | 2610.07704v1 |
Independent Multi-Agent Reinforcement Learning with Counterfactual Semantic-Social World Models |
| 2026-10-05 | 2610.06814v2 |
TAPDreamer: Transferable Adversarial Patches for World Action Models |
| 2026-10-05 | 2610.06349v1 |
KineWorld: Action-Induced Transport Fields for Embodied World Modeling |
| 2026-10-05 | 2610.06100v1 |
From Traces to Agentic Worlds: Agentic Language World Models for Interactive Environment Simulation |
| 2026-10-05 | 2610.05912v1 |
MiniCorp: The Last Mile of the AI Agent Firm |
| 2026-10-05 | 2610.05861v1 |
Imagine to Act: High-Fidelity Data Synthesis via Image Editing World Model for Scalable GUI Agent Training |
| 2026-10-04 | 2610.07028v1 |
Identifiable World Models from Pretrained Diffusion Representations |
| 2026-10-04 | 2610.05240v1 |
Pythia: Toward Foundation World Models for Multimodal Time Series |
abs:"world model" AND abs:"benchmark" — 命中 288，其中新增 17（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.10274v1 |
Sparse Planning in Visual World Models via Cost Gradients |
| 2026-10-07 | 2610.09546v1 |
PCDT: A Predictive Cognitive Digital Twin Framework for Intelligent and Autonomous 6G Network Ecosystems |
| 2026-10-07 | 2610.09514v1 |
STRIKE: Learning Visual State Transitions for Physical World Modeling |
| 2026-10-07 | 2610.09305v1 |
Kuration SDK: Addressing the Virtual2Real Gap via Data Curation |
| 2026-10-06 | 2610.08760v1 |
WorldSonus: Bringing Sound to Worlds |
| 2026-10-06 | 2610.08464v1 |
Federated Bayesian Surveillance of Mechanical Thrombectomy Adverse Events: A Population Risk Layer for Surgica |
| 2026-10-06 | 2610.07704v1 |
Independent Multi-Agent Reinforcement Learning with Counterfactual Semantic-Social World Models |
| 2026-10-05 | 2610.06814v2 |
TAPDreamer: Transferable Adversarial Patches for World Action Models |
| 2026-10-05 | 2610.05861v1 |
Imagine to Act: High-Fidelity Data Synthesis via Image Editing World Model for Scalable GUI Agent Training |
| 2026-10-03 | 2610.04475v1 |
VCLMU: Mechanism-Centric Virtual Cell World Modeling for Perturbation Response |
| 2026-10-03 | 2610.04301v1 |
EnvDreamer: Large-Scale Multimodal-to-Environment Generation for Embodied AI |
| 2026-10-03 | 2610.04168v1 |
Agentic Cognitive Depth: Operational Criteria for Evaluating LLM Agents |
| 2026-10-02 | 2610.03713v1 |
What Should World Models Forget? Stratified Retention for Continual Adaptation |
| 2026-10-02 | 2610.03356v1 |
ReFract: Benchmarking Perspective Awareness in Language Model Agents with Text World Models |
| 2026-10-02 | 2610.02957v1 |
Understanding Trajectory Heterogeneity in Federated World Model Learning |
| 2026-10-01 | 2610.02368v1 |
Rethinking World-Action Model for Compositional and In-Context Robotic Manipulation |
| 2026-10-01 | 2610.02331v1 |
World Editing: Intervening on Executable Worlds at Increasing Depth |
abs:"rollout" AND abs:"error accumulation" — 命中 40，其中新增 29（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-02 | 2610.03313v1 |
SDECast: Probabilistic Weather Forecasting in Continuous Time with Neural SDEs |
| 2026-10-02 | 2610.02779v1 |
TRAC: Trajectory-aware Reuse and Adaptive Correction for Efficient Autoregressive Video Generation |
| 2026-10-01 | 2610.01765v1 |
Physics-Refined Spatiotemporal Forecasting on Open-Boundary Hydrologic Graphs |
| 2026-09-29 | 2609.38114v1 |
Self-Aligned Forcing: Streaming Video Diffusion with Differentiable Noisy History |
| 2026-09-28 | 2609.34697v2 |
Triangular Resampling for Long-Horizon Motion Generation |
| 2026-09-27 | 2609.33412v1 |
Resolving State-Representation Mismatch: State-Space Visual Reasoning for Open-Loop VLA Planning |
| 2026-09-17 | 2609.19674v2 |
Conservation Buys Stability and Factoring Buys Counterfactuals in Physical World Models |
| 2026-09-15 | 2609.16621v1 |
Stable by Construction: Variational Latent Markov Operators for Long-Horizon PDE Prediction |
| 2026-09-08 | 2609.09123v1 |
Mask Forcing: Improving Autoregressive Video Diffusion Distillation via Dual-Noise Masking Rollout |
| 2026-09-03 | 2609.03225v2 |
Long-Horizon Consistent and Interaction-Aware World Models for Multi-Style End-to-End Driving |
| 2026-08-23 | 2608.22277v3 |
Dynamics-Aware Weighting for Deep Learning Forecasts of Chaotic Systems |
| 2026-08-16 | 2608.15815v1 |
KOALA: Koopman Operator Learning for WiFi-Based Anticipatory Hum |
| 2026-08-12 | 2608.12107v1 |
Avatar-Forever: Decoupled Parallel Training for High-Quality Real-Time Infinite Avatars |
| 2026-08-12 | 2608.11623v1 |
FM-LLM: A frequency-enhanced mixture-of-experts framework for adapting LLMs to time series forecasting |
| 2026-08-07 | 2608.07189v1 |
Autoregressive rollout error in latent-space reduced-order models of bluff-body wakes is accumulated phase dri |
| 2026-08-06 | 2608.06241v1 |
Timestep-Conditioned Transformers for Global Weather Forecasting |
| 2026-08-06 | 2608.05925v1 |
Local-Global Feature Mixer and Trend-Guided Consistent Learning for Remaining Useful Life Prediction of Rotati |
| 2026-08-06 | 2608.05806v1 |
Hierarchical Latent Prediction for Language Models |
| 2026-08-02 | 2608.01164v1 |
Hybrid Lagrangian-Eulerian Model for Lagrangian Fluid Simulation |
| 2026-07-31 | 2609.20309v1 |
Hypernetwork-Parameterized Spatially Adaptive Neural Operators for PDE Learning |
| 2026-07-31 | 2607.29135v1 |
HERO: History-Enriched Rollout Training for Long-Horizon Autoregressive Neural Operators |
| 2026-07-31 | 2608.11237v1 |
Geometry-aware Incremental Neural Operator for Long-Horizon PDE prediction |
| 2026-07-29 | 2607.27110v2 |
FreqForcing: Autoregressive Long Video Generation via Spectral Self-Anchoring |
| 2026-07-29 | 2607.27036v1 |
Mitigating Compounding Error via Video Representation Regularization |
| 2026-07-22 | 2607.19719v2 |
Koopman Dreamer: Spectrally Constrained Latent Dynamics for Stable World-Model Imagination |
| 2026-07-20 | 2607.18082v4 |
CriPO: Enhancing Rubric-based RL via Self-Distillation |
| 2026-07-17 | 2607.18309v1 |
Spatio-Temporal Prediction of Unsteady Airfoil Aerodynamics Using Augmented Graph Neural Ordinary Differential |
| 2026-07-13 | 2607.11971v1 |
Uncertainty-Aware Crack Growth Forecasting via Conditional Denoising Diffusion Models for Phase-Field Fracture |
| 2026-07-11 | 2607.10504v1 |
SUREFlow: State-space Uncertainty-aware REsidual Flow Matching for Robust Robot Manipulation |
abs:"video prediction" AND abs:"evaluation metrics" — 命中 0。复核过：词汇层面的假空白。同窗口换提法 —— abs:"video prediction" AND abs:"benchmark" 命中 14、abs:"video generation" AND abs:"evaluation metrics" 命中 6、abs:"video prediction" AND abs:"metrics" 命中 2。不限窗口时原提法本身也只有 3 条，且全部止于 2021 年。「evaluation metrics」是评测论文的写法，生成类论文写 benchmark / metric（单数）。不得记录为「视频预测没有评测工作」。
abs:"world model" AND abs:"diagnostic" — 命中 55，其中新增 23（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.09305v1 |
Kuration SDK: Addressing the Virtual2Real Gap via Data Curation |
| 2026-10-06 | 2610.08960v1 |
Directed Temporal Representations for Offline Visual Control |
| 2026-10-03 | 2610.04168v1 |
Agentic Cognitive Depth: Operational Criteria for Evaluating LLM Agents |
| 2026-09-30 | 2609.40003v1 |
DashVMC: Real-Time Discrete World Model Control in Geometry Dash |
| 2026-09-28 | 2609.36333v2 |
ATLAS: Aligned Transport of Latent Structure for Reliable World Model Planning |
| 2026-09-28 | 2609.35463v1 |
A.D.A.M.O. (Agent for language-Driven Actions with Multimodal Observations): A Visual-Symbolic Framework for V |
| 2026-09-26 | 2609.32679v1 |
The GUI Is Not the State: Diagnosing State Aliasing in GUI World Models |
| 2026-09-25 | 2609.30711v1 |
Recommendation World Models for Future-State Control |
| 2026-09-25 | 2609.30650v1 |
Causal Retention in Interactive Agents: Interface Factorization and Selective Adaptation |
| 2026-09-24 | 2609.30264v2 |
AD-WM: Action-Discriminative World Models for Counterfactual Model Predictive Control |
| 2026-09-23 | 2609.27621v2 |
SHRAV: State-Hypothesis-Reason-Action-Verify Framework for Physical Modeling and Inverse Design |
| 2026-09-14 | 2609.15781v1 |
When the World Lies: Backdoor Attacks on Latent World Models for Downstream Control |
| 2026-09-13 | 2609.14854v1 |
AutoLab: An Internet-Accessible Experimental Platform for Operational World Models in Wireless Networks |
| 2026-09-07 | 2609.07719v1 |
A radiographic world model for clinical reasoning and evidence generation |
| 2026-08-30 | 2608.29998v1 |
The Intervention Gap in Latent World Models |
| 2026-08-23 | 2608.22358v1 |
Tracing the Unlabeled Storm: Cross-Variable Transfer in a Lagrangian Atmospheric JEPA Framework |
| 2026-08-23 | 2608.22294v2 |
Beyond Instance Slots: Semantically Rich World Models for Physical Interaction Planning |
| 2026-08-19 | 2608.19085v2 |
DA-WAM: Decision-Aligned Future Latents for Driving World Models |
| 2026-08-17 | 2608.16859v2 |
HarnessEval-W: Agentifying the Evaluation of Visual Worlds |
| 2026-08-15 | 2608.15156v4 |
Low-Rank Dynamics-Effective Latent Carriers for Counterfactual Rollout in Learned World Models |
| 2026-08-13 | 2608.13049v1 |
H2R-Bench: Benchmarking Human-to-Robot Manipulation Video Generation in World Models |
| 2026-08-12 | 2609.05461v1 |
ARC-Bench: Closed-Loop Replanning Masks Broken Action Ranking in Frozen JEPA World Models |
| 2026-08-10 | 2608.10145v1 |
The Evaluation Protocol Determines the Result: An Independent Reproduction of LeWorldModel on TwoRoom |
abs:"world model" AND abs:"failure modes" — 命中 33，其中新增 26（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.10274v1 |
Sparse Planning in Visual World Models via Cost Gradients |
| 2026-10-05 | 2610.06582v1 |
Mind the Execution Gap: Action-Semantic Mismatch in World-Model Control |
| 2026-09-30 | 2609.39604v1 |
Why Do Conventional World Models Fail to Learn Cellular Automata? |
| 2026-09-28 | 2609.35463v1 |
A.D.A.M.O. (Agent for language-Driven Actions with Multimodal Observations): A Visual-Symbolic Framework for V |
| 2026-09-27 | 2609.33940v1 |
Behavioral Monitoring of JEPA World Models with Jacobian Centroids |
| 2026-09-26 | 2609.32761v1 |
From Feed-Forward to Flow: Unifying Reconstruction and Generation Is Easier Than You Think |
| 2026-09-26 | 2609.32679v1 |
The GUI Is Not the State: Diagnosing State Aliasing in GUI World Models |
| 2026-09-22 | 2609.26293v2 |
Dual-Frontier: When Can an Agent Trust Its World Model? |
| 2026-09-15 | 2609.17325v1 |
Intrinsic Motivation in Reinforcement Learning: A Research Agenda for Adaptive Self-Organisation |
| 2026-09-12 | 2609.13845v1 |
LePlanner: An Iterative Amortized Controller For World Models |
| 2026-09-05 | 2609.05834v1 |
Learning Counterfactual World Models for Embodied Reasoning under Partial Observability |
| 2026-09-02 | 2609.04264v3 |
Spectral-Target Physical Latent Structuring for JEPA-Style World Models |
| 2026-08-25 | 2608.26200v2 |
GameWAM: A World Action Model for Video Games |
| 2026-08-24 | 2608.23526v1 |
Correcting a learned physical invariant improves world-model rollouts |
| 2026-08-23 | 2608.22421v1 |
Where World Models Break: Natural-Input Failure Discovery |
| 2026-08-11 | 2608.10618v1 |
Toward the Cognitive--Physical Limits of Embodied Intelligence through a World-Model-Centric Autonomous Racing |
| 2026-08-06 | 2608.05720v2 |
PhyLatent: Learning Dynamics-Relevant Representations for JEPA World Models |
| 2026-07-29 | 2607.27511v1 |
Failure Detection for Surgical Robot Imitation Policies via Flow-Matching World Modeling |
| 2026-07-29 | 2607.26752v1 |
CalTwin: Towards Calibrated, Shift-Robust Medical World Models via Fisher-Information Regularisation |
| 2026-07-29 | 2607.26712v2 |
ActSWM: Action-Sensitive World Models for Long-Horizon Planning in Open-World Games |
| 2026-07-22 | 2607.19749v1 |
The World Model Remembers, the Actor Forgets: Dream Rehearsal for Continual Model-Based RL |
| 2026-07-17 | 2607.15620v1 |
AEGIS: Assay-Aware Protocol Validation and Runtime Monitoring for Open-Source Liquid Handling Robots |
| 2026-07-15 | 2607.13681v1 |
Towards Spatial Supersensing in the Wild |
| 2026-07-08 | 2607.07763v1 |
Unlocking Temporal Generalization in Hamiltonian Video Dynamics Models |
| 2026-07-06 | 2607.05352v2 |
Multiplayer Interactive World Models with Representation Autoencoders |
| 2026-06-26 | 2606.27806v3 |
Agent vs. Parametric World Models: Hybrid Planning for Reliable Language Agents |
08 · 联合嵌入预测与潜空间¶
abs:"JEPA" — 命中 266，其中新增 20（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.09940v1 |
Juno: Taming Predictive Latents for Vision-Language-Action Models |
| 2026-10-06 | 2610.09194v1 |
Patient, Place, Prior (P$^3$): What Counts as Personalization in Medical World Models? |
| 2026-10-06 | 2610.09048v1 |
Towards Financial World Modeling |
| 2026-10-06 | 2610.08400v1 |
Atom-JEPA: Joint-Embedding Predictive Architecture for 3D Atomistic Systems |
| 2026-10-06 | 2610.08192v1 |
MacJEPA: Missingness-Robust Audio-Visual Recognition from Untrimmed Egocentric Videos |
| 2026-10-05 | 2610.06008v1 |
Ultrasound Operator Guidance Using World Modeling and Retrieval Based Action Planning |
| 2026-10-05 | 2610.05996v1 |
EpicWorldModel: Exploration-driven Planning with Latent World Models |
| 2026-10-05 | 2610.05731v1 |
T-JEPA: A Temporal Joint-Embedding Predictive Architecture for Learning Better Remote Sensing Representations |
| 2026-10-04 | 2610.05409v1 |
BeliefGraph-JEPA: Structured Latent World Models for Action-Conditioned Time Series |
| 2026-10-04 | 2610.07025v1 |
WiSPER: Pose-Supervised Predictive and Residual Flow Refinement For Multi-Person 3D Pose Estimation With WiFi |
| 2026-10-04 | 2610.07006v1 |
STOCK-JEPA: Prior-Anchored Latent Revision Representation Learning in Equity Markets |
| 2026-10-04 | 2610.05043v1 |
CI-JEPA: A Counterfactual Analysis of Latent Representations in Joint-Embedding Predictive Architectures for S |
| 2026-10-03 | 2610.04778v1 |
Repeated-Measure Leakage, Distribution Shift, and Reliability under Partial Observation in Patient World Model |
| 2026-10-03 | 2610.06965v1 |
ACG-WAM: World-Action Modeling via Action-Conditioned Geometric Latent Prediction |
| 2026-10-02 | 2610.03374v1 |
EVEWorld: Physical Evolution Supervision for Embodied World Models |
| 2026-10-02 | 2610.03106v1 |
S2S-JEPA: Predicting the Predictable at Subseasonal-to-Seasonal Timescales |
| 2026-10-02 | 2610.02864v1 |
NeuroLens: Learning Latent Embeddings of Neural Semantics from Chronic Recordings |
| 2026-10-01 | 2610.01947v1 |
Latent JEPA: Abstract Future Prediction for Latent Reasoning in Chemistry |
| 2026-10-01 | 2610.00976v2 |
Variational Streaming Flow: Probabilistic Forecasting in Physical Time |
| 2026-09-30 | 2610.00722v1 |
JEPA-TTT: Persistent Test-Time Training of Latent World Models for Planning under Dynamics Shifts |
abs:"joint embedding" AND abs:"prediction" — 命中 152，其中新增 20（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.09940v1 |
Juno: Taming Predictive Latents for Vision-Language-Action Models |
| 2026-10-06 | 2610.08400v1 |
Atom-JEPA: Joint-Embedding Predictive Architecture for 3D Atomistic Systems |
| 2026-10-05 | 2610.05996v1 |
EpicWorldModel: Exploration-driven Planning with Latent World Models |
| 2026-10-05 | 2610.05731v1 |
T-JEPA: A Temporal Joint-Embedding Predictive Architecture for Learning Better Remote Sensing Representations |
| 2026-10-04 | 2610.05543v1 |
AngularWM: Wireless World Modeling for AoA Prediction from Multi-Antenna I/Q |
| 2026-10-04 | 2610.05409v1 |
BeliefGraph-JEPA: Structured Latent World Models for Action-Conditioned Time Series |
| 2026-10-04 | 2610.05240v1 |
Pythia: Toward Foundation World Models for Multimodal Time Series |
| 2026-10-04 | 2610.07006v1 |
STOCK-JEPA: Prior-Anchored Latent Revision Representation Learning in Equity Markets |
| 2026-10-04 | 2610.05043v1 |
CI-JEPA: A Counterfactual Analysis of Latent Representations in Joint-Embedding Predictive Architectures for S |
| 2026-10-02 | 2610.03106v1 |
S2S-JEPA: Predicting the Predictable at Subseasonal-to-Seasonal Timescales |
| 2026-10-02 | 2610.02864v1 |
NeuroLens: Learning Latent Embeddings of Neural Semantics from Chronic Recordings |
| 2026-10-01 | 2610.01947v1 |
Latent JEPA: Abstract Future Prediction for Latent Reasoning in Chemistry |
| 2026-10-01 | 2610.00976v2 |
Variational Streaming Flow: Probabilistic Forecasting in Physical Time |
| 2026-09-30 | 2610.00722v1 |
JEPA-TTT: Persistent Test-Time Training of Latent World Models for Planning under Dynamics Shifts |
| 2026-09-30 | 2609.40129v1 |
VR-JEPA: Learning Contrastive-State Latent Guidance for Generation-based Video Reasoning |
| 2026-09-29 | 2609.36952v1 |
ER-JEPA: Experience Replay Improves Joint-Embedding Predictive Learning in Language Models |
| 2026-09-28 | 2609.35603v2 |
Control-Geometry Straightening for Sampling-Based Latent Planning |
| 2026-09-28 | 2609.34407v1 |
Beyond Textual Chain-of-Thought: JEPA-Conditioned Latent Reasoning for Large Audio Language Models |
| 2026-09-28 | 2609.34085v1 |
AD-E2E-JEPA: A Joint-Embedding Predictive Architecture For End-to-End Autonomous Driving |
| 2026-09-27 | 2609.33698v1 |
One Latent, Many Tokens: Jointly Learning Compressed Embeddings for Efficient Language Diffusion |
abs:"representation collapse" AND abs:"world model" — 命中 6，其中新增 2
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-09-28 | 2609.36305v1 |
Bilinear World Models: Learning Representations with Structured Dynamics for Efficient Control |
| 2026-09-02 | 2609.04264v3 |
Spectral-Target Physical Latent Structuring for JEPA-Style World Models |
abs:"latent space" AND abs:"world model" — 命中 83，其中新增 22（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-05 | 2610.06008v1 |
Ultrasound Operator Guidance Using World Modeling and Retrieval Based Action Planning |
| 2026-10-02 | 2610.04009v1 |
SUAVE: Unified Video-Action Models via Masked Diffusion |
| 2026-10-01 | 2610.01942v1 |
Latent-Foresight: End-to-End Learning Predictable Representations for Latent World Models |
| 2026-10-01 | 2610.01842v1 |
On the Divergence of Accuracy and Mechanism Consistency in Time Series World Models |
| 2026-10-01 | 2610.01224v1 |
Supervise What Decides Success: Criterion-Aligned Auxiliary Losses for Latent World-Model Planning |
| 2026-09-30 | 2609.38927v1 |
World-as-Graph: Relational World Modeling Through Latent Space Graphs |
| 2026-09-29 | 2609.36985v1 |
Abductive World Modeling via Causal Representation Learning |
| 2026-09-29 | 2609.36845v1 |
DSWM: Decomposed Spatio-Temporal World Model for Demand-Driven UAV Base Station Repositioning |
| 2026-09-28 | 2609.34375v1 |
LRC-JEPA: Disentangling Dynamics and Residual Context for Efficient World Models |
| 2026-09-28 | 2609.34300v1 |
When World Models Lie: Adaptive Safety Analysis Under Wrong Imaginations |
| 2026-09-28 | 2609.34206v1 |
WorldGuide: Learning Success-Failure Boundaries in Latent World Models for Vision-Language-Action Policies |
| 2026-09-25 | 2609.31162v1 |
WorldTS: World Modeling for Multimodal Covariate-aware Time Series Forecasting |
| 2026-09-24 | 2609.30436v1 |
WALT: Learning World-Model-Aligned Latent Trajectories for Autonomous Driving |
| 2026-09-24 | 2609.30214v1 |
Underwater C3-JEPA: An Object-Centric Cross-View World Model for ROV Salvage |
| 2026-09-23 | 2609.28414v1 |
Frozen Flows Forget: Diagnosing and Restoring Lost Motion in a Latent-flow World Model |
| 2026-09-21 | 2609.24048v1 |
What Matters in Designing World Action Models: An Empirical Study |
| 2026-09-18 | 2609.22521v2 |
Latent Policy Steering: An Efficient and Flexible Framework for Cross-Embodiment Transfer |
| 2026-09-12 | 2609.13845v1 |
LePlanner: An Iterative Amortized Controller For World Models |
| 2026-09-11 | 2609.12874v1 |
VideoTok4D: A 4D-Aware Video Tokenizer for Compact World Representation |
| 2026-09-11 | 2609.12347v1 |
DWMP: Leveraging Dual World Models for Humanoid Obstacle Traversal |
| 2026-09-02 | 2609.04264v3 |
Spectral-Target Physical Latent Structuring for JEPA-Style World Models |
| 2026-08-29 | 2608.28995v2 |
Hydra: A Navigation World Action Model with Discrete Latent Planning and Continuous Flow-Matching Execution |
abs:"energy-based" AND abs:"world model" — 命中 0。复核过：低频真线，只是这 120 天恰好空窗。不限窗口命中 8，且集中在 2026 上半年：2605.07199（Three-in-One World Model，能量一致性）、2602.23058（GeoWorld）。同族提法 abs:"EBM" AND abs:"world model" 不限窗口命中 1（2406.08862 Cognitively Inspired Energy-Based World Models）。对照：abs:"energy-based model" 不限窗口 630、同窗口 35，说明词本身不冷，是「energy-based + world model」这个组合少。可以写「这一线稀疏」，不得写「没有」。
abs:"self-supervised" AND abs:"world model" — 命中 36，其中新增 27（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-04 | 2610.05543v1 |
AngularWM: Wireless World Modeling for AoA Prediction from Multi-Antenna I/Q |
| 2026-10-01 | 2610.02515v1 |
IGNITE Tokamak World Model Architecture |
| 2026-10-01 | 2610.01201v1 |
iSEE: Object Permanence Through Self-Supervision |
| 2026-09-30 | 2610.00722v1 |
JEPA-TTT: Persistent Test-Time Training of Latent World Models for Planning under Dynamics Shifts |
| 2026-09-30 | 2609.39182v1 |
MEND: Label-Free Detection, Localisation, and Correction of Latent Hallucination in World Models |
| 2026-09-29 | 2609.38278v1 |
Masked Swingers: Harnessing Data Augmentation to Advance Autoencoders for Self-Supervised Learning |
| 2026-09-28 | 2609.34375v1 |
LRC-JEPA: Disentangling Dynamics and Residual Context for Efficient World Models |
| 2026-09-28 | 2609.34085v1 |
AD-E2E-JEPA: A Joint-Embedding Predictive Architecture For End-to-End Autonomous Driving |
| 2026-09-25 | 2609.32013v2 |
TriO: Tri-Modal Unsupervised Occupancy World Model for Anything Perception |
| 2026-09-25 | 2609.30667v1 |
StarWM: Self-Supervised Trained Attention Routing for Robust World Models |
| 2026-09-23 | 2609.28414v1 |
Frozen Flows Forget: Diagnosing and Restoring Lost Motion in a Latent-flow World Model |
| 2026-09-23 | 2609.28049v1 |
Prompt, Probe, Train, or Annotate? Single-camera sports video understanding in amateur settings |
| 2026-09-22 | 2609.25541v1 |
A JEPA Recipe for Tabular Foundation Models |
| 2026-09-18 | 2609.21740v1 |
Sandwich-Residuals: Parameter-Efficient Test-time Adaptation of World Models |
| 2026-09-09 | 2609.09627v1 |
Seven Sources of Physical AI Capability Formation |
| 2026-08-30 | 2608.29998v1 |
The Intervention Gap in Latent World Models |
| 2026-08-11 | 2608.11174v2 |
VIScore: Diagnosing Planning-Relevant Quality in Latent World Models |
| 2026-08-11 | 2608.11026v1 |
MAJEPPA: Morphing and Assessing in a Unified Piano Performance Space |
| 2026-08-10 | 2608.09771v1 |
SLIM-0.5B: Learning Action-Grounded Predictive Latents for Robot Manipulation |
| 2026-08-10 | 2609.26118v1 |
GDLAM: Group-Disentangled Latent Action Model for Highly Disentangled Embodied Pretraining |
| 2026-08-07 | 2608.07409v1 |
UniJEPA: A Unified Joint-Embedding Predictive Architecture for Task-Agnostic Visual World Modeling |
| 2026-08-06 | 2608.05706v1 |
LAWM-3D: Learning 3D-Aware Latent Actions from Human Videos for Generalizable Robot World Models |
| 2026-08-05 | 2608.04378v1 |
Helping Music Co-Creation Agents 'Listen' Well: Hierarchical Self-Supervised World Models for Understanding an |
| 2026-07-30 | 2607.28624v1 |
PhiZero: A World Model Built Around Physical Language |
| 2026-07-24 | 2607.22000v1 |
Music-JEPA: Learning a World Model of Sound from Action |
| 2026-07-05 | 2607.04500v1 |
Geographic Diversity Beats Data Volume for Cross-Domain Generalization in Zero-Label JEPA Driving World Models |
| 2026-07-03 | 2607.03198v2 |
Reduced-Order Models: The Mother of World Models |
09 · 长时程与交互式生成¶
abs:"long-horizon" AND abs:"world model" — 命中 179，其中新增 17（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-06 | 2610.08640v1 |
RIWANav: Recursive World-Action Models with Self-Improvement for Urban Navigation |
| 2026-10-05 | 2610.06637v1 |
Long-Horizon Textual World Modeling through Structured Reasoning |
| 2026-10-05 | 2610.06100v1 |
From Traces to Agentic Worlds: Agentic Language World Models for Interactive Environment Simulation |
| 2026-10-04 | 2610.04916v1 |
PreAct-Nav: Agentic Reasoning Before Action for Urban Navigation |
| 2026-10-02 | 2610.04009v1 |
SUAVE: Unified Video-Action Models via Masked Diffusion |
| 2026-10-01 | 2610.02368v1 |
Rethinking World-Action Model for Compositional and In-Context Robotic Manipulation |
| 2026-10-01 | 2610.00976v2 |
Variational Streaming Flow: Probabilistic Forecasting in Physical Time |
| 2026-09-30 | 2609.40003v1 |
DashVMC: Real-Time Discrete World Model Control in Geometry Dash |
| 2026-09-30 | 2609.39727v1 |
OverForge: Reasoning Through Strategies and Tactics Helps Cooperative Lifelong Adaptation |
| 2026-09-30 | 2609.39101v1 |
Beyond Prediction: Steering VLM Agents with Retrospective World Modeling |
| 2026-09-30 | 2609.38839v1 |
FrameMorrow: Future-guided Frame Selection with Prospective Tokens for Long-Horizon Video Generation |
| 2026-09-29 | 2609.38562v1 |
LongTake: Learning to Sustain Dynamics in Long-Horizon Video Generation |
| 2026-09-29 | 2609.36851v2 |
RoXDrive: Closed-Loop Reinforcement Learning for End-to-End Autonomous Driving via Action-Faithful Rollouts |
| 2026-09-28 | 2609.36344v1 |
DeepRewind: Predicting and Repairing Premature Commitments in Deep Research Agents |
| 2026-09-28 | 2609.35375v1 |
From Pixel to Poses: Object-centric Tool Manipulation Learning from Human Demonstrations |
| 2026-09-28 | 2609.35138v2 |
FlexiWorld: Learning and Planning via Flexible Action Chunks Across Multiple Time Scales |
| 2026-09-28 | 2609.34470v1 |
Precise Editing and Flexible Referencing for Interactable Worlds |
abs:"interactive world model" — 命中 35，其中新增 21（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.09785v1 |
UltraWorld: Learning Interactive Ultrasound World Models from Untracked Clinical Videos with Acoustic Sampling |
| 2026-10-06 | 2610.08941v1 |
SPW-Nav: A Streaming Panoramic World Model for Language-Guided Navigation |
| 2026-10-04 | 2610.04920v1 |
PWM: Personalized World Models with Online Reinforcement Learning |
| 2026-10-01 | 2610.02331v1 |
World Editing: Intervening on Executable Worlds at Increasing Depth |
| 2026-09-29 | 2609.38123v1 |
HelixWorld: A Real-time Interactive Audio-Visual World Model |
| 2026-09-28 | 2609.34144v1 |
CAST: Reconstruction-Coupled Acceleration of Interactive World Models |
| 2026-09-25 | 2609.31893v1 |
CyberWorld: World Models for Sample-Efficient Autonomous Cyber Defense |
| 2026-09-07 | 2609.07532v1 |
PhysReal: Learning Real-World Deformable Object Physics via Hybrid Constitutive Modeling |
| 2026-09-07 | 2609.07051v1 |
TrojanWorld: Backdooring World-Model Agents via Imagination Steering |
| 2026-09-01 | 2609.01560v1 |
H3-World: Turning Language Understanding into World Control |
| 2026-08-26 | 2608.25479v3 |
4DStreamCtrl: Interactive Video Generation with Online 4D Control |
| 2026-08-18 | 2608.21439v1 |
WorldMind: Decoupled Game World Model for State-Aware NPC Behavior |
| 2026-08-14 | 2608.14022v1 |
ForgeWM: Progressive Causal Training for Few-Step Action-Conditioned Video World Models |
| 2026-08-13 | 2608.13546v2 |
Alaya-EVOKE: From Linear-Scaling Supervision to Endless World |
| 2026-08-10 | 2608.09449v2 |
Sekai2: From World Exploration to Interactive World Modeling |
| 2026-07-30 | 2607.28624v1 |
PhiZero: A World Model Built Around Physical Language |
| 2026-07-30 | 2607.28362v1 |
ShadowDancer: Teaching Video World Models Any Action by Learning Unified Dynamics Representations from a Video |
| 2026-07-23 | 2607.21594v1 |
Streaming Multi-Agent Autoregressive Diffusion Model with World State Registers |
| 2026-07-21 | 2607.18703v1 |
Generative World Renderer at the Speed of Play |
| 2026-07-11 | 2607.10389v1 |
Stateful Worlds, Stateless Elasticity: Exact-State Serving for Interactive World Models |
| 2026-07-07 | 2607.06216v2 |
MoWorld: A Flash World Model |
abs:"action-conditioned" AND abs:"video generation" — 命中 24，其中新增 21
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-09-30 | 2609.40153v1 |
Dream4ACT: A Shared Visual Action Interface for Multi-Embodiment Video-Action Modeling |
| 2026-09-30 | 2609.38839v1 |
FrameMorrow: Future-guided Frame Selection with Prospective Tokens for Long-Horizon Video Generation |
| 2026-09-29 | 2609.38059v1 |
WorldLine: Action-Driven Visual Simulation for Robotic Manipulation |
| 2026-09-23 | 2609.28811v1 |
DeltaWAM: Delta World Action Models for Bimanual Manipulation |
| 2026-08-31 | 2608.30897v1 |
CAER: Causal Action Effect Reweighting for World Model Training |
| 2026-08-27 | 2608.27406v2 |
CLAP: Cross-Embodiment Video World Models are Zero-Shot Physical Simulators |
| 2026-08-25 | 2608.24199v2 |
NVIDIA Cosmos-H-Dreams: Real-Time Generative Physics Simulation for Surgical Robotics |
| 2026-08-23 | 2608.22403v1 |
LD4WAM: Learning Latent Dynamics from Human Videos for World Action Models |
| 2026-08-18 | 2608.18077v1 |
Hydra-0: Action Flow for Generalist World Modeling and Control |
| 2026-08-14 | 2608.14022v1 |
ForgeWM: Progressive Causal Training for Few-Step Action-Conditioned Video World Models |
| 2026-07-31 | 2607.29302v1 |
BWM: A Low-Cost High-Fidelity World Simulator for Robot Learning |
| 2026-07-29 | 2607.26579v1 |
ContactFlow: A video action conditioning that transfers across embodiments |
| 2026-07-07 | 2607.06018v1 |
RoboTALES: Learning Reasoning-Guided Robot Policies via Task-Aligned Simulated Futures |
| 2026-06-28 | 2606.29501v1 |
Learning Transferable Dynamics Priors from Action to World Modeling |
| 2026-06-24 | 2606.26410v2 |
Neural Voxel Dynamics: Learning Volumetric Feature Advection for 3D Physics in V-JEPA Latent Space |
| 2026-06-24 | 2606.25473v1 |
Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation i |
| 2026-06-22 | 2606.23296v1 |
IOI: Decoupling Kinematics and Physics for Interactive World Models |
| 2026-06-17 | 2606.18610v3 |
SC3-Eval: Evaluating Robot Foundation Models via Self-Consistent Video Generation |
| 2026-06-16 | 2606.18180v1 |
EgoCS-400K: An Egocentric Gameplay Dataset for World Models |
| 2026-06-14 | 2606.15768v1 |
LaWAM: Latent World Action Models for Efficient Dynamics-Aware Robot Policies |
| 2026-06-13 | 2606.15341v1 |
CausalDrive: Real-time Causal World Models for Autonomous Driving |
abs:"object permanence" — 命中 13，其中新增 13
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-02 | 2610.03713v1 |
What Should World Models Forget? Stratified Retention for Continual Adaptation |
| 2026-10-01 | 2610.01201v1 |
iSEE: Object Permanence Through Self-Supervision |
| 2026-09-28 | 2609.34554v1 |
Where Memory Belongs: Ledger, an Object Ledger for Memory-Augmented VLAs |
| 2026-09-23 | 2609.28654v1 |
Training Object Permanence in World Models |
| 2026-09-17 | 2609.20819v2 |
Can 4D Foundation Models Remember? |
| 2026-08-27 | 2608.26794v1 |
Ring Forcing: Towards Precise Long-Term Memory for Autoregressive Video Diffusion |
| 2026-08-03 | 2608.02289v1 |
Extended Field of View Analysis for VideoGAN-based Trajectory Generation |
| 2026-08-03 | 2608.01614v1 |
Linear Multi-Timescale Retention as a Memory-Efficient Vision-Language Bridge |
| 2026-07-14 | 2607.12231v1 |
The GEST-Engine: From Event Graphs to Synthetic Video. A Full Technical Report |
| 2026-07-10 | 2607.09138v1 |
BeyondSight: Object Permanence for End-to-End Autonomous Driving |
| 2026-06-29 | 2606.30481v1 |
Situation Perception: A Necessary Primitive to Artificial Superintelligence |
| 2026-06-26 | 2606.28455v3 |
Event-Conditioned Diagnostics of Kinematic, Contact, and Object-Permanence Structure in Passive Object-State W |
| 2026-06-15 | 2606.20707v1 |
GEOPHYS: The Geometry of Physical Plausibility |
abs:"autoregressive" AND abs:"video prediction" — 命中 4，其中新增 3
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-06 | 2610.09090v1 |
Tucker Bottleneck Attention for Multi-Dimensional Sequence Modeling |
| 2026-09-16 | 2609.18430v1 |
StrucPhysVideo: Learning Physical Dynamics from Structured Captions and Robot Actions |
| 2026-07-14 | 2607.13031v1 |
The Seriality Gap in Video Diffusion Models |
abs:"memory" AND abs:"long-horizon" AND abs:"navigation" — 命中 41，其中新增 26（只取了最近 30 条，新增数是下界）
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-04 | 2610.05024v1 |
LightVLN: Efficient Aerial Vision-and-Language Navigation with Compact Memory and History-Guided Local Aggrega |
| 2026-10-04 | 2610.04916v1 |
PreAct-Nav: Agentic Reasoning Before Action for Urban Navigation |
| 2026-09-30 | 2609.39915v1 |
NavHarness: Adaptive Goals for Agentic Vision-Language Navigation |
| 2026-09-28 | 2609.35431v1 |
Memory in the Sky: Low-Altitude Question Answering with Multi-Agent Memory Aggregation |
| 2026-09-28 | 2609.34163v1 |
Reliability-Aware Sparse Route Memory for Round-Trip Vision-Language Navigation |
| 2026-09-26 | 2609.32626v1 |
World SLAM Model: Joint World Modeling for SLAM and Navigation |
| 2026-09-23 | 2609.27612v1 |
RegenHarness: A Robot Agent Harness with Evidence-Gated Recursive Self-Improvement |
| 2026-09-22 | 2609.25666v1 |
Deploying Foundation Models for Embodied Navigation |
| 2026-09-20 | 2609.23465v1 |
Propose, Verify, Commit: Evidence-Grounded Memory for Long-Horizon Multi-Actor Conversations |
| 2026-09-08 | 2609.08442v1 |
AirAnchor: Bridging Local and Global Spatial Information for Zero-Shot Aerial Vision-and-Language Navigation |
| 2026-09-08 | 2609.08159v1 |
OmniNav: Robust Long-Horizon Target Navigation in Dynamic Environments |
| 2026-08-31 | 2608.30396v1 |
Scaffolding Foundation Models into Physical-World Agents Pushes the Frontier of Long-Horizon Navigation |
| 2026-08-30 | 2608.29596v2 |
A Systematic Survey of Agentic Skills: Architecture, Lifecycle, and Security |
| 2026-08-29 | 2608.29114v1 |
CGFM-Nav: Cognitive Graph-Field Memory for Semantic-Guided Lifelong Multimodal Embodied Navigation |
| 2026-08-26 | 2609.29555v1 |
Visual Representation and History Modeling for Navigation World Models |
| 2026-08-24 | 2608.23383v2 |
Long-Horizon Audio-Visual Generation for Persistent Stories and Interactive Worlds |
| 2026-08-21 | 2608.21690v1 |
Context as an Environment: Programmatic Context Management for Long-Horizon Agents |
| 2026-08-04 | 2608.05013v1 |
OneDayAgent: Towards a Long-Horizon Harness for Autonomous Agents |
| 2026-08-02 | 2608.01456v1 |
Long-Horizon Embodied Decision-Making via Multimodal Memory Compression |
| 2026-07-31 | 2607.29600v1 |
HAM-VLN: Harnessing Hierarchical Agentic Memory for Zero-Shot Vision-and-Language Navigation |
| 2026-07-26 | 2607.23504v1 |
MemVLN: Episodic and Procedural Memory for Vision-and-Language Navigation |
| 2026-07-20 | 2607.17599v1 |
ConsiSpace: Learning Geometric Consistency Matters for Video Spatial Reasoning |
| 2026-07-19 | 2607.17038v1 |
Reward-Driven LLM Agent Workflows: Synthesizing POMDP Routing and Self-Correction for Autonomous Decision-Maki |
| 2026-07-14 | 2607.12370v1 |
StratMamba: Strategic and Reactive Stream Partitioning for Path-Efficient LiDAR-Based Obstacle Avoidance |
| 2026-07-13 | 2607.11377v1 |
A Glimpse into Long-term Physical Coexistence with Intelligent Robots |
| 2026-07-11 | 2607.10350v3 |
ABot-AgentOS: A General Robotic Agent OS with Lifelong Multi-modal Memory |
uav · 无人机主线¶
abs:"UAV" AND abs:"world model" — 命中 17，其中新增 14
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.09335v1 |
SearchWorld: Spatial Value-Grounded Imagination for UAV Object Search via World Models |
| 2026-09-29 | 2609.36845v1 |
DSWM: Decomposed Spatio-Temporal World Model for Demand-Driven UAV Base Station Repositioning |
| 2026-09-20 | 2609.23656v1 |
WOLF: World Model Guided LiDAR Exploration with Predictive Frontiers |
| 2026-09-16 | 2609.18326v1 |
UAVs Meet Embodied Intelligence: Bridging Human Intents and Flying Dynamics Via Harnessing Physical-Digital AI |
| 2026-08-31 | 2609.00106v1 |
Deploying and Evaluating a Smart-Agriculture Agentic Engine for Full-Season Soybean Farm Operations |
| 2026-08-20 | 2608.20126v1 |
RMWorld: Task-Aware Radio World Models with Value-of-Information Guided Multi-Trial Learning for Multi-UAV Com |
| 2026-08-06 | 2608.05792v1 |
When Agentic AI Meets Integrated Sensing and Communication |
| 2026-08-06 | 2608.05597v1 |
Uncertainty-Aware World Model for Aerial Image-Goal Navigation |
| 2026-07-30 | 2607.27865v1 |
Learning to Understand Body Language from Flight through Robust 3D Avatar Placing |
| 2026-07-28 | 2607.25728v1 |
Shared Voxel-Map-Based Cooperative Indoor UAV Guidance with a Multi-Agent Soft Actor-Critic Controller |
| 2026-07-22 | 2607.19719v2 |
Koopman Dreamer: Spectrally Constrained Latent Dynamics for Stable World-Model Imagination |
| 2026-07-10 | 2607.09078v1 |
Toward Active Object Detection for UAVs in the Wild: A Large-Scale Dataset, Benchmark and Method |
| 2026-07-07 | 2607.06706v1 |
Vision Language Action (VLA) Models for Unmanned Aerial Robotics and Bimanual Manipulation: A Review |
| 2026-07-05 | 2607.04352v1 |
Last-Meter Precision Navigation for UAVs: A Diffusion-Refined Aerial Visual Servoing Approach |
abs:"aerial" AND abs:"world model" — 命中 18，其中新增 14
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-10-07 | 2610.09335v1 |
SearchWorld: Spatial Value-Grounded Imagination for UAV Object Search via World Models |
| 2026-09-29 | 2609.36845v1 |
DSWM: Decomposed Spatio-Temporal World Model for Demand-Driven UAV Base Station Repositioning |
| 2026-09-23 | 2609.27621v2 |
SHRAV: State-Hypothesis-Reason-Action-Verify Framework for Physical Modeling and Inverse Design |
| 2026-09-20 | 2609.23656v1 |
WOLF: World Model Guided LiDAR Exploration with Predictive Frontiers |
| 2026-09-16 | 2609.18326v1 |
UAVs Meet Embodied Intelligence: Bridging Human Intents and Flying Dynamics Via Harnessing Physical-Digital AI |
| 2026-08-20 | 2608.20126v1 |
RMWorld: Task-Aware Radio World Models with Value-of-Information Guided Multi-Trial Learning for Multi-UAV Com |
| 2026-08-06 | 2608.05792v1 |
When Agentic AI Meets Integrated Sensing and Communication |
| 2026-08-06 | 2608.05597v1 |
Uncertainty-Aware World Model for Aerial Image-Goal Navigation |
| 2026-08-02 | 2608.01049v1 |
FactorJEPA: Factorizing Monolithic Futures into Layout-Agent-Interaction Channels for Crowded and Chaotic Glob |
| 2026-07-30 | 2607.27865v1 |
Learning to Understand Body Language from Flight through Robust 3D Avatar Placing |
| 2026-07-10 | 2607.09661v1 |
PanoWorld: Real-World Panoramic Generation |
| 2026-07-10 | 2607.09078v1 |
Toward Active Object Detection for UAVs in the Wild: A Large-Scale Dataset, Benchmark and Method |
| 2026-07-07 | 2607.06706v1 |
Vision Language Action (VLA) Models for Unmanned Aerial Robotics and Bimanual Manipulation: A Review |
| 2026-07-05 | 2607.04352v1 |
Last-Meter Precision Navigation for UAVs: A Diffusion-Refined Aerial Visual Servoing Approach |
abs:"drone" AND abs:"world model" — 命中 7，其中新增 6
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-08-31 | 2609.00106v1 |
Deploying and Evaluating a Smart-Agriculture Agentic Engine for Full-Season Soybean Farm Operations |
| 2026-07-30 | 2607.27865v1 |
Learning to Understand Body Language from Flight through Robust 3D Avatar Placing |
| 2026-07-28 | 2607.25728v1 |
Shared Voxel-Map-Based Cooperative Indoor UAV Guidance with a Multi-Agent Soft Actor-Critic Controller |
| 2026-07-20 | 2607.17669v2 |
Attention from Above: A Multimodal Model for Drone-Based Object Localization |
| 2026-07-07 | 2607.06706v1 |
Vision Language Action (VLA) Models for Unmanned Aerial Robotics and Bimanual Manipulation: A Review |
| 2026-06-23 | 2606.24152v2 |
Autonomous Video Generation with Counterfactual Controllability for Self-Evolving World Models |
abs:"aerial" AND abs:"video generation" — 命中 0。复核过：纯粹是窗口假象。不限窗口命中 6，最近三条就在窗口外几周：2605.19728（Aero-World，2026-05-19）、2605.15964（WorldVLN，2026-05-15）、2604.07991（MotionScape，2026-04-09）。同窗口换提法 abs:"drone" AND abs:"video generation" 命中 1（2606.24152）、abs:"UAV" AND abs:"video generation" 命中 1（2610.02451）。这是本轮最该记的一条：窗口起点卡在 6 月 10 日，把 2026 年 4–5 月空中世界模型那一批整批切掉了——写「近期空白」之前必须先看不限窗口的命中。
abs:"UAV" AND abs:"video prediction" — 命中 0。复核过：同上是窗口假象。不限窗口命中 2：2606.06147（WorldFly，2026-06-04，比窗口起点早 6 天）、2512.21710（RAPTOR，2025-12-25）。同窗口换提法 abs:"aerial" AND abs:"video prediction"、abs:"aerial" AND abs:"future frame prediction" 也都命中 0。不得写成「无人机不做视频预测」。
abs:"aerial" AND abs:"navigation" AND abs:"world model" — 命中 5，其中新增 4
| 日期 | arXiv | 标题 |
|---|---|---|
| 2026-09-16 | 2609.18326v1 |
UAVs Meet Embodied Intelligence: Bridging Human Intents and Flying Dynamics Via Harnessing Physical-Digital AI |
| 2026-08-06 | 2608.05597v1 |
Uncertainty-Aware World Model for Aerial Image-Goal Navigation |
| 2026-07-07 | 2607.06706v1 |
Vision Language Action (VLA) Models for Unmanned Aerial Robotics and Bimanual Manipulation: A Review |
| 2026-07-05 | 2607.04352v1 |
Last-Meter Precision Navigation for UAVs: A Diffusion-Refined Aerial Visual Servoing Approach |
本文件由 tools/watch.py 生成，重跑即可刷新。收录进正文前，ID 与标题一律以 arXiv API 为准。