This paper introduces a multi-degree-of-freedom reinforcement learning framework for robotic 3D measurement, enabling continuous viewpoint planning to improve the reconstruction of complex geometries. The framework uses a voxel-based state representation with dynamic ray-traced coverage updates and a dual-objective reward function to balance overlap control and viewpoint minimization. Experimental results on industrial parts show the proposed method achieves superior overlap regulation and planning efficiency compared to existing techniques, leading to more accurate 3D reconstructions.

This paper introduces a novel control framework that combines conformal prediction (CP) and system level synthesis (SLS) to achieve robust out-of-distribution (OOD) planning and control with learned dynamics models. The method uses weighted CP with a learned covariance model to derive high-confidence model error bounds, which are then incorporated into an SLS-based robust nonlinear MPC formulation with volume-optimized reachable sets for constraint tightening. Empirical results on nonlinear systems like a 4D car and a 12D quadcopter demonstrate improved safety and robustness, particularly in OOD scenarios, compared to baselines.

The paper introduces Incremental Signature Contribution (ISC), a method that decomposes truncated path signatures into a temporally ordered sequence of elements, preserving the algebraic structure and expressivity of signatures while exposing temporal evolution. This allows for processing signature-based representations using sequential models, addressing the limitation of standard path signatures which collapse temporal structure. The authors then introduce ISC-Transformer (ISCT), an offline RL model integrating ISC into a standard Transformer, and demonstrate its effectiveness on benchmark tasks, particularly in settings requiring temporal sensitivity.

The paper introduces Affordance-Graphed Task Worlds (AGT-World), a framework that automatically generates interactive simulated environments and robot task policies from real-world observations by formalizing the task space as a structured graph. This graph-based approach allows for hierarchical decomposition of complex goals into atomic primitives, addressing the limitations of random proposal or static replication methods. The authors further incorporate a self-evolution mechanism with hybrid feedback, combining Vision-Language Model reasoning and geometric verification, to refine policies.

This paper introduces a task planning framework that integrates Learning-Informed Object Search (LIOS) actions into high-level planning to address scenarios with missing objects. The framework models LIOS actions as deterministic, leveraging model-based calculations to estimate their cost and interleave search and execution steps. The approach demonstrates effective task planning with uncertainty, outperforming both non-learned and learned baselines in simulated ProcTHOR environments and real-world experiments involving retrieval and meal preparation tasks.

This paper introduces VLAW, an iterative algorithm for co-improving vision-language-action (VLA) policies and action-conditioned video generation world models using real-world rollouts. VLAW leverages real-world data to refine the world model, which is then used to generate synthetic data for further policy improvement, addressing the limitations of world models trained solely on demonstration datasets. Experiments on a real robot demonstrate a 39.2% absolute improvement in success rate over the base policy, highlighting the effectiveness of the iterative co-improvement strategy.

This paper analyzes the InvestESG multi-agent simulation to characterize conditions leading to intertemporal social dilemmas where individual incentives conflict with collective welfare. It then applies Advantage Alignment, an opponent shaping algorithm, to influence agent learning within InvestESG, demonstrating its ability to systematically favor socially beneficial equilibria. The work provides theoretical justification for why Advantage Alignment promotes cooperation and shows that shaping agent learning can improve outcomes related to sustainability goals.

The paper introduces 3DGSNav, a zero-shot object navigation (ZSON) framework that leverages 3D Gaussian Splatting (3DGS) as persistent memory for vision-language models (VLMs) to improve spatial reasoning. 3DGSNav actively constructs a 3DGS representation of the environment and uses trajectory-guided free-viewpoint rendering to generate frontier-aware first-person views, which are then combined with structured visual prompts and Chain-of-Thought prompting to enhance VLM reasoning. Experiments on multiple benchmarks and a quadruped robot show that 3DGSNav achieves competitive performance compared to existing methods.

The paper introduces a framework for intelligent AI delegation, enabling AI agents to decompose complex tasks and delegate sub-components to other AI agents or humans. This framework addresses limitations in current task decomposition methods by incorporating elements like authority transfer, accountability, and trust-building. The authors propose an adaptive approach applicable to both AI and human agents within complex delegation networks, contributing to the development of protocols for agentic systems.

The paper introduces V-SHiNE, a browser-based virtual smart home environment designed to facilitate the evaluation of explainable AI (XAI) methods in the context of smart home automation. V-SHiNE enables researchers to configure realistic smart home environments, simulate user behaviors, integrate custom explanation engines, and log user interactions. A user study with 159 participants demonstrates the framework's utility for assessing the impact and quality of different explanation strategies.

The paper introduces DTAPP-IICR, a Delivery-Time Aware Prioritized Planning method with Incremental and Iterative Conflict Resolution, for preflight planning of large UAV fleets in dynamic airspaces with temporal No-Fly Zones and heterogeneous vehicle profiles. DTAPP-IICR uses a novel 4D single-agent planner (SFIPP-ST) to generate roundtrip trajectories while enforcing temporal NFZs and modeling inter-agent conflicts as soft constraints, followed by a Large Neighborhood Search guided by a geometric conflict graph. Experiments on benchmarks with up to 1,000 UAVs demonstrate near-100% success and up to 50% runtime reduction compared to batch Enhanced Conflict-Based Search, showcasing its scalability and practicality for dense urban airspace.

The paper identifies limitations in current Vision-Language-Action (VLA) models stemming from inadequate visual representations learned through language-image contrastive learning or image-based self-supervised learning. It proposes JEPA-VLA, a method that integrates video predictive embeddings (specifically V-JEPA 2) into VLAs to improve environment understanding and policy priors. Experiments on benchmarks like LIBERO and real-robot tasks demonstrate that JEPA-VLA significantly improves performance by leveraging the ability of video predictive embeddings to encode task-relevant temporal dynamics.

The paper introduces LDA-1B, a robot foundation model that scales to 1B parameters by learning dynamics, policy, and visual forecasting from a new 30k-hour embodied interaction dataset (EI-30k) comprising diverse human and robot trajectories. LDA-1B leverages a structured DINO latent space for dynamics prediction to avoid pixel-space modeling and employs a multi-modal diffusion transformer to handle asynchronous vision and action streams. Experimental results demonstrate that LDA-1B outperforms existing methods on contact-rich, dexterous, and long-horizon tasks, while also enabling data-efficient fine-tuning by effectively utilizing low-quality trajectories.

The paper introduces Code2Worlds, a framework for generating 4D dynamic scenes by formulating the task as language-to-simulation code generation. It addresses the challenges of multi-scale context entanglement and the semantic-physical execution gap by using a dual-stream architecture for disentangled object and environment generation, combined with a physics-aware closed-loop mechanism involving a PostProcess Agent and VLM-Motion Critic. Experiments on the Code4D benchmark demonstrate that Code2Worlds significantly outperforms existing methods in scene generation score (SGS) and richness, while also generating more physically plausible dynamics.

The paper introduces HAIC, a framework for humanoid robots to interact with underactuated objects having independent dynamics, addressing limitations of prior HOI methods focused on rigidly coupled objects. HAIC uses a dynamics predictor to estimate high-order object states from proprioceptive history, projecting these onto geometric priors to create a dynamic occupancy map for collision avoidance and contact affordance inference. Through asymmetric fine-tuning of a world model, HAIC achieves robust performance on agile manipulation tasks like skateboarding and cart pushing, as well as long-horizon multi-object tasks.

This paper introduces Counterfactual Conditional Likelihood (CCL) rewards to address redundant exploration in multiagent systems by scoring each agent's unique contribution to team exploration. CCL rewards agents for observations that are informative with respect to the joint exploration of the team, rather than solely for individual novelty. Experiments in continuous multiagent domains demonstrate that CCL accelerates learning in sparse reward environments requiring tight coordination.

The paper introduces GigaBrain-0.5M*, a vision-language-action (VLA) model trained using world model-based reinforcement learning to improve multi-step action prediction. They leverage the spatiotemporal reasoning capabilities of video world models pre-trained on large video datasets to enhance VLA learning. By integrating world model-based reinforcement learning via RAMP (Reinforcement leArning via world Model-conditioned Policy), GigaBrain-0.5M* achieves significant performance gains (approximately 30%) over the RECAP baseline on complex manipulation tasks and demonstrates reliable long-horizon execution in real-world deployments.

This paper introduces a method for learning structured latent representations in RL where distances reflect transition costs, providing a geometric interpretation of uncertainty without explicit probabilistic modeling. They achieve this with a multimodal latent transition model and inverse distance weighting for sensor fusion, enabling adaptive integration of multiple sensor modalities. Empirical validation on multimodal RL tasks demonstrates improved robustness to sensor noise, superior state estimation, and enhanced RL agent performance compared to baselines, eliminating the need for noise augmentation.

This paper introduces Adaptive-RF Transmission (ART), a communication-aware planning algorithm for multi-agent robotic exploration that modulates transmission location based on signal strength and data payload size. ART aims to improve coordination and efficiency in communication-limited environments by enabling heterogeneous robot teams to share information without excessive backtracking. Simulation results across cave-inspired environments show that ART and its extension, ART-SST, outperform existing strategies, achieving significant reductions in distance traveled and exploration time.

This paper introduces Adaptive-Horizon Conflict-Based Search (ACCBS), a closed-loop multi-agent path finding algorithm that addresses the limitations of open-loop planners and closed-loop heuristics in MAPF. ACCBS employs a finite-horizon CBS variant with a horizon-changing mechanism inspired by iterative deepening MPC, dynamically adjusting the planning horizon based on computational budget. The algorithm reuses a single constraint tree to enable seamless transitions between horizons, achieving anytime behavior and asymptotic optimality.

The paper investigates whether neural world models truly learn physical laws or rely on statistical shortcuts, particularly under out-of-distribution shifts. They introduce PhyIP, a non-invasive evaluation protocol that assesses the linear decodability of physical quantities from frozen latent representations, contrasting it with adaptation-based methods. Their results show that when self-supervised learning achieves low error, latent physical structures are linearly accessible and robust to OOD shifts, while adaptation-based evaluations can collapse this structure, suggesting that non-invasive probes are more accurate for evaluating physical world models.

This paper connects Joint-Embedding Predictive Architectures (JEPAs) with Quasimetric Reinforcement Learning (QRL) by focusing on a specific class of JEPA energy functions: intrinsic (least-action) energies defined as infima of accumulated local effort. It demonstrates that under closure and additivity assumptions, intrinsic energies are quasimetrics, aligning JEPAs trained on these energies with the quasimetric value functions used in QRL for goal-reaching control. The work highlights the structural mismatch between symmetric energies and one-way reachability, advocating for asymmetric (quasimetric) energies in scenarios where directionality is important.

This paper investigates the impact of bit allocation strategies on the performance of world model-based planning, specifically using DINO-WM on the Wall planning task. The study compares uniform, mixed, asymmetric, and layerwise quantization schemes under different planner budgets to identify critical bitwidth thresholds. Results show a sensitivity to bit allocation in the 4-bit regime, with encoder precision being particularly important for maintaining performance, suggesting the need for module-aware quantization policies.

The paper introduces Multi-Graph Search (MGS), a novel search-based motion planning algorithm for high-dimensional robotic systems that addresses the limitations of existing methods in terms of motion consistency and computational cost. MGS maintains and expands multiple implicit graphs, focusing exploration on promising regions and merging disconnected subgraphs as needed. The authors prove completeness and bounded suboptimality of MGS and demonstrate its effectiveness on manipulation and mobile manipulation tasks.

This paper introduces a modular multi-LLM pipeline for generating agricultural simulation environments in Unreal Engine from natural language prompts, addressing limitations of existing LLM-based 3D scene generation approaches. The pipeline incorporates 3D asset retrieval, domain knowledge injection, and code generation, enhanced by LLM optimization techniques like few-shot prompting, RAG, and finetuning. Experiments demonstrate the system's effectiveness in creating realistic and semantically accurate agricultural environments, offering significant time savings compared to manual design.

The paper introduces Any House Any Task (AHAT), a household task planner designed for long-horizon planning in large environments with ambiguous instructions. AHAT trains an LLM to map task instructions and textual scene graphs into PDDL subgoals, which are then solved using symbolic reasoning for optimal plan generation. To improve decomposition of complex intentions, they propose TGPO, a reinforcement learning algorithm integrating external correction of intermediate reasoning traces into Group Relative Policy Optimization (GRPO), leading to significant performance gains.

The paper introduces ReaDy-Go, a real-to-sim pipeline that generates photorealistic dynamic scenarios using 3D Gaussian Splatting (GS) to train visual navigation policies robust to the sim-to-real gap and moving obstacles. ReaDy-Go combines a static GS scene with dynamic human GS avatars driven by plausible motions derived from 2D trajectories, and uses a robot expert planner designed for dynamic GS representations to generate navigation datasets. Experiments demonstrate that policies trained with ReaDy-Go outperform baselines in both simulation and real-world environments, exhibiting improved navigation performance and generalization.

The paper introduces WorldTree, a novel framework for dynamic scene reconstruction from monocular video that addresses limitations in spatiotemporal decomposition. WorldTree uses a Temporal Partition Tree (TPT) for coarse-to-fine temporal optimization and Spatial Ancestral Chains (SAC) for hierarchical spatial dynamics. Experiments demonstrate that WorldTree achieves state-of-the-art performance, improving LPIPS by 8.26% on NVIDIA-LS and mLPIPS by 9.09% on DyCheck compared to existing methods.

This paper introduces INTENT, a novel inference-time planning framework for budget-constrained, tool-augmented LLMs that addresses the challenge of costly tool use in sequential decision-making. INTENT uses an intention-aware hierarchical world model to anticipate future tool usage and risk-calibrated costs, enabling more effective online decision-making. Experiments on a cost-augmented StableToolBench demonstrate that INTENT achieves superior task success while strictly adhering to budget constraints, even under dynamic market conditions.

The paper introduces GRXForm, a Graph Transformer model for amortized molecular optimization that sequentially adds atoms and bonds to a molecule. To improve generalization, the authors identify and address the high variance in rewards caused by heterogeneous starting structures by using Group Relative Policy Optimization (GRPO). GRXForm demonstrates strong generalization to out-of-distribution molecular scaffolds, achieving competitive performance with instance optimizers in multi-objective optimization without requiring inference-time oracle calls or refinement.

The paper introduces Trajectory-Search Rollouts (TSR), a training-time method that uses lightweight tree search to improve the quality of rollouts in multi-turn reinforcement learning for LLM agents. TSR selects high-scoring actions at each turn during rollout generation using task-specific feedback, leading to more informative training trajectories. Experiments on Sokoban, FrozenLake, and WebShop demonstrate that TSR, when combined with PPO and GRPO, achieves up to 15% performance gains and more stable learning.

The paper introduces Gaia2, a benchmark designed to evaluate LLM agents in dynamic, asynchronous environments where the environment evolves independently of agent actions. Gaia2 features scenarios requiring agents to handle temporal constraints, adapt to noisy events, resolve ambiguity, and collaborate, coupled with write-action verifiers for fine-grained evaluation. Evaluations of state-of-the-art models reveal trade-offs between reasoning, efficiency, and robustness, with GPT-5 achieving the highest overall score (42% pass@1) but struggling with time-sensitive tasks.

The paper introduces RISE, a robotic reinforcement learning framework that leverages a Compositional World Model to improve policy learning in simulation. This world model predicts multi-view futures using a controllable dynamics model and evaluates outcomes with a progress value model, generating advantages for policy improvement. By training the policy entirely in a closed-loop, self-improving imaginary environment, RISE achieves significant performance gains over existing methods in dynamic manipulation tasks.

The authors introduce MolmoSpaces, a large-scale, open-source ecosystem comprising over 230k diverse indoor environments and 130k richly annotated object assets, designed to address the limitations of existing robot benchmarks in capturing the long tail of real-world scenarios. This simulator-agnostic ecosystem supports a wide range of embodied tasks, including navigation, manipulation, and long-horizon planning, and includes MolmoSpaces-Bench, a benchmark suite of 8 tasks. Experiments demonstrate strong sim-to-real correlation and highlight sensitivities to factors like prompt phrasing and camera occlusion, establishing MolmoSpaces as a valuable resource for scalable robot learning research.

This paper introduces $\chi_{0}$, a resource-efficient framework designed to enhance the robustness of long-horizon robotic manipulation by addressing distributional inconsistencies between demonstration, policy learning, and execution. The framework incorporates model arithmetic for merging diverse demonstration distributions, stage-aware advantage estimation for stable progress signals, and train-deploy alignment techniques using augmentation and DAgger. Experimental results demonstrate that $\chi_{0}$ significantly outperforms existing methods in garment manipulation tasks, achieving a 250% improvement in success rate compared to $\pi_{0.5}$ with limited data and compute resources.

The paper introduces Dreaming in Code (DiCode), a framework that uses foundation models to generate executable environment code variations for curriculum learning in open-ended environments. DiCode addresses the challenge of discovering learnable sequences of experiences in complex environments by "dreaming" code-level variations of the world to scaffold learning. Experiments in the Craftax environment demonstrate that DiCode enables agents to acquire long-horizon skills, achieving a 16% improvement in mean return over the strongest baseline and success on late-game combat tasks where prior methods fail.

The paper introduces HiCrowd, a hierarchical framework combining reinforcement learning (RL) and model predictive control (MPC) to improve robot navigation in dense crowds. A high-level RL policy selects a "follow point" to align the robot with compatible crowd flows, while a low-level MPC tracks this point with short-horizon planning for safety. Experiments on real-world and synthetic datasets demonstrate that HiCrowd outperforms reactive and learning-based baselines in navigation efficiency, safety, and reducing freezing behaviors.

This paper introduces a Digital Twin-enabled access scheduling framework inspired by the Dual Mind World Model (DMWM) architecture for intelligent network control in dynamic environments. The DMWM framework combines short-horizon predictive planning with symbolic model-based rollout, allowing the scheduler to anticipate network states and optimize transmission decisions. Experimental results demonstrate that DMWM outperforms traditional heuristics and reinforcement learning baselines in bursty, interference-limited, and deadline-sensitive environments, while also improving interpretability and sample efficiency.

This survey paper reviews the evolution of end-to-end autonomous driving, contrasting traditional modular systems with emerging end-to-end approaches based on imitation learning, reinforcement learning, and foundation models. It categorizes recent advances in planning, reasoning, data generation, and scene understanding within the context of large language models (LLMs) and vision-language models (VLMs). The paper also addresses challenges like multimodal fusion complexity and safety risks, and outlines future research directions including world models and modular foundation model architectures.

The paper introduces NetWorld, a Communication-based Diffusion World Model, to improve few-shot generalization across heterogeneous MARL tasks in wireless networks. NetWorld pre-trains a classifier-guided conditional diffusion model on multi-task offline datasets and performs trajectory planning within the learned world model, avoiding online interaction. The model incorporates a mean-field communication mechanism to address non-stationarity and promote coordination.

This paper introduces a dynamic weighted spherical particle swarm optimization (DW-SPSO) algorithm to address the challenges of UAV path planning in complex environments, specifically the issues of high-dimensional search spaces, local optima, and real-world constraints. DW-SPSO employs a dual Sigmoid-based adaptive weight adjustment mechanism to balance exploration and exploitation, along with lens-based opposition learning to enhance search flexibility. Experiments on real digital elevation models demonstrate that DW-SPSO outperforms state-of-the-art PSO variants regarding path safety, smoothness, and convergence speed, validated by the Wilcoxon signed-rank test.

This paper introduces a novel memory replay model (MPR) inspired by hippocampal-prefrontal collaboration in the human brain, aiming to improve robot navigation in complex environments. The model integrates meta-reinforcement learning with model-based policy optimization, creating a closed-loop interaction between hippocampal planning replay and prefrontal policy evaluation networks. Experimental results demonstrate that MPR achieves a 14% reduction in navigation path length and a 10% higher success rate compared to other models, along with significantly improved learning efficiency.

The paper introduces HumanDiffusion, an image-conditioned diffusion-based trajectory planner for UAVs in search and rescue, enabling navigation towards humans detected via YOLO-V3. By predicting trajectories directly in pixel space from RGB images, the system avoids reliance on pre-existing maps or computationally expensive planning. Experiments in simulation and real-world scenarios demonstrate a mean squared error of 0.02 in pixel-space trajectory reconstruction and an 80% mission success rate, highlighting the method's effectiveness for human-aware navigation.

The paper introduces a "Just-in-Time" (JiT) framework for simulation-based reasoning, positing that humans construct simplified environment representations online to overcome computational limitations. The JiT model interleaves simulation, visual search, and representation modification, using ongoing simulations to guide visual attention and identify relevant objects for encoding. Empirical validation in grid-world planning and physical reasoning tasks demonstrates the model's ability to make accurate predictions while encoding only a small subset of objects, supporting the idea that humans create reduced representations for efficient mental simulation.

This paper introduces an Active Inference framework for decentralized UAV swarm trajectory design, aiming to improve adaptability and safety in dynamic environments. A hierarchical World Model is trained on expert trajectories generated by a Genetic Algorithm with Repulsion Forces (GA-RF) to represent swarm behavior at different levels of abstraction. The UAVs then use active inference to select actions that minimize the divergence between their beliefs and the World Model's predictions.

This paper introduces DreamWaQ++, a multimodal reinforcement learning framework that fuses proprioceptive and exteroceptive information for robust quadrupedal locomotion in complex environments. The approach trains a controller capable of agile navigation across challenging terrains like rough ground, steep slopes, and high stairs, while also exhibiting resilience to out-of-distribution scenarios. Key to the success is the fusion of proprioceptive feedback with exteroceptive data to enable obstacle avoidance and adaptive gait planning.

This paper introduces a multifaceted approach to accelerate reservoir simulation by combining advanced software techniques, AI/ML-based enhancements, and scalable hardware solutions. The software innovations include multiscale SFI methods for black-oil models and AI/ML for phase labeling and saturation pressure prediction in compositional models, while hardware strategies involve CPU, GPU, and cloud-based execution. Applied to real-world, high-resolution reservoirs, the combined approach achieved up to 4x runtime reduction using multiscale SFI, up to 2x speedups with full-GPU execution, and up to 4x improvements with AI/ML, significantly compressing field development planning timelines.

The paper introduces an integrated path planning and trajectory tracking system for autonomous vehicles, combining an improved A* algorithm for global planning, a fuzzy dynamic window approach (DWA) for local planning, and a fuzzy PID controller for trajectory tracking. The improved A* enhances search efficiency and path quality through an enhanced heuristic function, redundant node removal, and path smoothing. The fuzzy DWA enables smooth obstacle avoidance, and the fuzzy PID controller adaptively adjusts parameters for precise path following. Real-world experiments on a ROS-based vehicle demonstrate superior performance compared to conventional approaches in path planning efficiency, obstacle avoidance, and path smoothness.

This paper introduces LOONG, a novel planning and control framework for time-optimal MAV flight in cluttered environments. The framework combines imitation learning to accelerate time allocation for polynomial trajectory generation with a time-optimal model predictive contouring control (MPCC) that incorporates safe flight corridor (SFC) constraints. Experimental results on a LiDAR-based MAV platform demonstrate superior aggressiveness and a peak speed of 18 m/s in real-world cluttered environments, showcasing the framework's robustness.

The paper introduces SPICE, a Bayesian in-context reinforcement learning (ICRL) method that learns a prior over Q-values using a deep ensemble and updates it at test-time via Bayesian updates using in-context information. This approach addresses limitations of existing ICRL methods by enabling improvement beyond the training distribution and robustness to suboptimal training data. SPICE achieves regret-optimal behavior in stochastic bandits and finite-horizon MDPs, even when pre-trained on suboptimal trajectories, and demonstrates superior empirical performance compared to existing ICRL and meta-RL methods on bandit and control benchmarks.