Visual observation of objects is essential for many robotic applications, such as object reconstruction and manipulation, navigation, and scene understanding. Machine learning algorithms constitute the state-of-the-art in many fields but require vast data sets, which are costly and time-intensive to collect. Automated strategies for observation and exploration are crucial to enhance the efficiency of data gathering. Therefore, a novel strategy utilizing the Next-Best-Trajectory principle is developed for a robot manipulator operating in dynamic environments. Local trajectories are generated to maximize the information gained from observations along the path while avoiding collisions. We employ a voxel map for environment modeling and utilize raycasting from perspectives around a point of interest to estimate the information gain. A global ergodic trajectory planner provides an optional reference trajectory to the local planner, improving exploration and helping to avoid local minima. To enhance computational efficiency, raycasting for estimating the information gain in the environment is executed in parallel on the graphics processing unit. Benchmark results confirm the efficiency of the parallelization, while real-world experiments demonstrate the strategy's effectiveness.
Objective: Cardiotoxicity is a devastating complication of thoracic radiotherapy. Current radiotherapy imaging protocols are insufficient to decouple and quantify cardiac motion, limiting substructure-specific motion considerations in treatment planning. We propose a 5D-MRI workflow for substructure-specific motion analysis, with future extension to margin calculation. Approach: Our 5D-MRI workflow was implemented for 10 healthy volunteers, ranging from 23 to 65 years old, reconstructing images for end-exhale/inhale and active-exhale/inhale for end-systole/diastole. For motion assessment, proximal coronary arteries, chambers, great vessels, and cardiac valves/nodes were contoured across all images and verified.Centroid/bounding box excursion was calculated for cardiac, respiratory, and hysteresis motion. Distance metrics were tested for statistical independence across substructure pairings. Main Results: 5D-MRI images were successfully acquired and contoured for all volunteers. Cardiac motion was greatest for the coronary arteries (specifically the right coronary) and smallest for the great vessels. Respiratory motion was dominant in the S-I direction and largest for the inferior vena cava. Respiratory hysteresis was generally <5 mm but exceeded 5 mm for some volunteers. For cardiac motion, there were statistical differences between the coronary arteries, chambers, and great vessels, and between the right/left heart. Respiratory motion differed significantly between the base and apex of the heart. Significance: Our 5D-MRI workflow successfully decouples cardiorespiratory motion with one ~5-minute acquisition. Cardiac motion was >5mm for the coronary arteries and chambers, while respiratory motion was >5mm for all substructures. Statistical considerations and inter-patient variability indicate a substructure and patient-specific approach may be needed for PRV assessment.
Active mobility is becoming an essential component of the green transition in modern cities. However, the challenge of designing an efficient network of protected bike lanes without disrupting existing road networks for motorised vehicles remains unsolved. This paper focuses on the specific case of Milan, using a network approach that considers street widths to optimise the placement of dedicated bike lanes at the edges of the network. Unlike other network approaches in this field, our method considers the actual shapes of the streets, which introduces a realistic aspect lacking in current studies. We used these data to simulate cycling networks that maximise connectivity while minimising the impact of bike lane placement on the drivable network. Our mixed simulation strategies optimise for edge betweenness and width. Furthermore, we quantify the impact of dedicated bike lane infrastructure on the existing road network, demonstrating that it is feasible to create highly effective cycling networks with minimal disruption caused by lane width reductions. This paper illustrates how realistic cycling lanes can be simulated using road width data and discusses the challenges and benefits of moving beyond one-dimensional road data in network studies.
Nature has inspired humans in different ways. The formation behavior of animals can perform tasks that exceed individual capability. For example, army ants could transverse gaps by forming bridges, and fishes could group up to protect themselves from predators. The pattern formation task is essential in a multiagent robotic system because it usually serves as the initial configuration of downstream tasks, such as collective manipulation and adaptation to various environments. The formation of complex shapes, especially hollow shapes, remains an open question. Traditional approaches either require global coordinates for each robot or are prone to failure when attempting to close the hole due to accumulated localization errors. Inspired by the ribbon idea introduced in the additive self-assembly algorithm by the Kilobot team, we develop a two-stage algorithm that does not require global coordinates information and effectively forms shapes with holes. In this paper, we investigate the partitioning of the shape using ribbons in a hexagonal lattice setting and propose the add-subtract algorithm based on the movement sequence induced by the ribbon structure. This advancement opens the door to tasks requiring complex pattern formations, such as the assembly of nanobots for medical applications involving intricate structures and the deployment of robots along the boundaries of areas of interest. We also provide simulation results on complex shapes, an analysis of the robustness as well as a proof of correctness of the proposed algorithm.
We introduce Scenario Dreamer, a fully data-driven generative simulator for autonomous vehicle planning that generates both the initial traffic scene - comprising a lane graph and agent bounding boxes - and closed-loop agent behaviours. Existing methods for generating driving simulation environments encode the initial traffic scene as a rasterized image and, as such, require parameter-heavy networks that perform unnecessary computation due to many empty pixels in the rasterized scene. Moreover, we find that existing methods that employ rule-based agent behaviours lack diversity and realism. Scenario Dreamer instead employs a novel vectorized latent diffusion model for initial scene generation that directly operates on the vectorized scene elements and an autoregressive Transformer for data-driven agent behaviour simulation. Scenario Dreamer additionally supports scene extrapolation via diffusion inpainting, enabling the generation of unbounded simulation environments. Extensive experiments show that Scenario Dreamer outperforms existing generative simulators in realism and efficiency: the vectorized scene-generation base model achieves superior generation quality with around 2x fewer parameters, 6x lower generation latency, and 10x fewer GPU training hours compared to the strongest baseline. We confirm its practical utility by showing that reinforcement learning planning agents are more challenged in Scenario Dreamer environments than traditional non-generative simulation environments, especially on long and adversarial driving environments.
Emergency preparedness reduces the severity and impact of major disasters. In the case of earthquakes, a rapid and efficient emergency response is essential to reduce the number of fatalities. Therefore, the design and planning of an adequate emergency transportation network are crucial in earthquake-prone locations. In the context of emergency transportation modeling, the aim of emergency routing is to find the network with the minimum length that can provide access between the maximum number of Emergency Response Centers (ERCs) and damaged areas. Meanwhile, the goal of the facility location and allocation problem is to optimize the placement of temporary hospitals to increase coverage and accessibility, particularly in remote or severely impacted areas. This paper proposes a multi-objective, robust, multi-modal, and multi-time-period optimization problem that simultaneously optimizes routing, facility location, and hospital allocation. The objective function is to minimize unmet commodity demand, unserved injuries, and economic costs. We adopt a fuzzy goal programming approach to solve the multi-objective simultaneous routing, facility location, and allocation model.
Residential electricity demand forecasting is critical for efficient energy management and grid stability. Accurate predictions enable utility companies to optimize planning and operations. However, real-world residential electricity demand data often exhibit intricate temporal variability, including multiple seasonalities, periodicities, and abrupt fluctuations, which pose significant challenges for forecasting models. Previous models that rely on statistical methods, recurrent, convolutional neural networks, and transformers often struggle to capture these intricate temporal dynamics. To address these challenges, we propose the Seasonal-Periodic Decomposition Network (SPDNet), a novel deep learning framework consisting of two main modules. The first is the Seasonal-Trend Decomposition Module (STDM), which decomposes the input data into trend, seasonal, and residual components. The second is the Periodical Decomposition Module (PDM), which employs the Fast Fourier Transform to identify the dominant periods. For each dominant period, 1D input data is reshaped into a 2D tensor, where rows represent periods and columns correspond to frequencies. The 2D representations are then processed through three submodules: a 1D convolution to capture sharp fluctuations, a transformer-based encoder to model global patterns, and a 2D convolution to capture interactions between periods. Extensive experiments conducted on real-world residential electricity load data demonstrate that SPDNet outperforms traditional and advanced models in both forecasting accuracy and computational efficiency. The code is available in this repository: https://github.com/Tims2D/SPDNet.
Charged-lepton flavor violation (cLFV) is one of the most powerful probes for New Physics (NP). Since lepton flavor conservation is an accidental symmetry in the Standard Model (SM), it is naturally violated in many NP models, with contributions at the level of the current experimental sensitivities. Moreover, the negligible SM contributions would make the observation of cLFV unambiguous evidence of NP. It makes these searches extremely sensitive and, at the same time, extremely pure. Thanks to the intense muon beams currently available, their intriguing upgrade programs, and the progress in the detection techniques, cLFV muon processes are the golden channels in this field. Experimental programs to search for $\mu^+ \to e^+ \gamma$, $\mu^+ \to e^+ e^+ e^-$ and the $\mu \to e$ conversion in the nuclear field are currently ongoing. We review the current status and the strategic plans for future searches. This document is an update of the prior cLFV submission to the 2018 European Strategy for Particle Physics (ESPP); the earlier submission should be consulted for more experimental details.
This work presents a scalable Bayesian modeling framework for evaluating building energy performance using smart-meter data from 2,788 Danish single-family homes. The framework leverages Bayesian statistical inference integrated with Energy Signature (ES) models to characterize thermal performance in buildings. This approach quantifies key parameters such as the Heat Loss Coefficient (HLC), solar gain, and wind infiltration, while providing full posterior distributions to reflect parameter uncertainty. Three model variants are developed: a baseline ES model, an auto-regressive model (ARX-ES) to account for thermal inertia, and an auto-regressive moving average model (ARMAX-ES) that approximates stochastic gray-box dynamics. Results show that model complexity improves one-step-ahead predictive performance, with the ARMAX-ES model achieving a median Bayesian R^2 of 0.94 across the building stock. At the single-building level, the Bayesian approach yields credible intervals for yearly energy demand within $\pm1\%$, enabling more robust diagnostics than deterministic methods. Beyond improved accuracy, the Bayesian framework enhances decision-making by explicitly representing uncertainty in building performance parameters. This provides a more realistic foundation for investment prioritization, demand forecasting, and long-term energy planning. The method is readily applicable to other building typologies or geographies, offering a scalable tool for data-driven energy management under uncertainty.
Accurate classification of building roof types from aerial imagery is crucial for various remote sensing applications, including urban planning, disaster management, and infrastructure monitoring. However, this task is often hindered by the limited availability of labeled data for supervised learning approaches. To address this challenge, this paper investigates the effectiveness of self supervised learning with EfficientNet architectures, known for their computational efficiency, for building roof type classification. We propose a novel framework that incorporates a Convolutional Block Attention Module (CBAM) to enhance the feature extraction capabilities of EfficientNet. Furthermore, we explore the benefits of pretraining on a domain-specific dataset, the Aerial Image Dataset (AID), compared to ImageNet pretraining. Our experimental results demonstrate the superiority of our approach. Employing Simple Framework for Contrastive Learning of Visual Representations (SimCLR) with EfficientNet-B3 and CBAM achieves a 95.5% accuracy on our validation set, matching the performance of state-of-the-art transformer-based models while utilizing significantly fewer parameters. We also provide a comprehensive evaluation on two challenging test sets, demonstrating the generalization capability of our method. Notably, our findings highlight the effectiveness of domain-specific pretraining, consistently leading to higher accuracy compared to models pretrained on the generic ImageNet dataset. Our work establishes EfficientNet based self-supervised learning as a computationally efficient and highly effective approach for building roof type classification, particularly beneficial in scenarios with limited labeled data.
Precisely grasping an object is a challenging task due to pose uncertainties. Conventional methods have used cameras and fixtures to reduce object uncertainty. They are effective but require intensive preparation, such as designing jigs based on the object geometry and calibrating cameras with high-precision tools fabricated using lasers. In this study, we propose a method to reduce the uncertainty of the position and orientation of a grasped object without using a fixture or a camera. Our method is based on the concept that the flat finger pads of a parallel gripper can reduce uncertainty along its opening/closing direction through flat surface contact. Three orthogonal grasps by parallel grippers with flat finger pads collectively constrain an object's position and orientation to a unique state. Guided by the concepts, we develop a regrasp planning and admittance control approach that sequentially finds and leverages three orthogonal grasps of two robotic arms to eliminate uncertainties in the object pose. We evaluated the proposed method on different initial object uncertainties and verified that the method has satisfactory repeatability accuracy. It outperforms an AR marker detection method implemented using cameras and laser jet printers under standard laboratory conditions.
Multi-Agent Pathfinding is used in areas including multi-robot formations, warehouse logistics, and intelligent vehicles. However, many environments are incomplete or frequently change, making it difficult for standard centralized planning or pure reinforcement learning to maintain both global solution quality and local flexibility. This paper introduces a hybrid framework that integrates D* Lite global search with multi-agent reinforcement learning, using a switching mechanism and a freeze-prevention strategy to handle dynamic conditions and crowded settings. We evaluate the framework in the discrete POGEMA environment and compare it with baseline methods. Experimental outcomes indicate that the proposed framework substantially improves success rate, collision rate, and path efficiency. The model is further tested on the EyeSim platform, where it maintains feasible Pathfinding under frequent changes and large-scale robot deployments.
Automated construction is one of the most promising areas that can improve efficiency, reduce costs and minimize errors in the process of building construction. In this paper, a comparative analysis of three neural network models for semantic segmentation, U-Net(light), LinkNet and PSPNet, is performed. Two specialized datasets with images of houses built from wooden cubes were created for the experiments. The first dataset contains 4 classes (background, foundation, walls, roof ) and is designed for basic model evaluation, while the second dataset includes 44 classes where each cube is labeled as a separate object. The models were trained with the same hyperparameters and their accuracy was evaluated using MeanIoU and F1 Score metrics. According to the results obtained, U-Net(light) showed the best performance with 78% MeanIoU and 87% F1 Score on the first dataset and 17% and 25% respectively on the second dataset. The poor results on the second dataset are due to the limited amount of data, the complexity of the partitioning and the imbalance of classes, making it difficult to accurately select individual cubes. In addition, overtraining was observed in all experiments, manifested by high accuracy on the training dataset and its significant decrease on the validation dataset. The present work is the basis for the development of algorithms for automatic generation of staged building plans, which can be further scaled to design complete buildings. Future research is planned to extend the datasets and apply methods to combat overfitting (L1/L2 regularization, Early Stopping). The next stage of work will be the development of algorithms for automatic generation of a step-by-step plan for building houses from cubes using manipulators. Index Terms-Deep Learning, Computer vision, CNN, Semantic segmentation, Construction materials.
Vision-language models (VLMs) have demonstrated remarkable capabilities in robotic planning, particularly for long-horizon tasks that require a holistic understanding of the environment for task decomposition. Existing methods typically rely on prior environmental knowledge or carefully designed task-specific prompts, making them struggle with dynamic scene changes or unexpected task conditions, e.g., a robot attempting to put a carrot in the microwave but finds the door was closed. Such challenges underscore two critical issues: adaptability and efficiency. To address them, in this work, we propose an adaptive multi-agent planning framework, termed REMAC, that enables efficient, scene-agnostic multi-robot long-horizon task planning and execution through continuous reflection and self-evolution. REMAC incorporates two key modules: a self-reflection module performing pre-condition and post-condition checks in the loop to evaluate progress and refine plans, and a self-evolvement module dynamically adapting plans based on scene-specific reasoning. It offers several appealing benefits: 1) Robots can initially explore and reason about the environment without complex prompt design. 2) Robots can keep reflecting on potential planning errors and adapting the plan based on task-specific insights. 3) After iterations, a robot can call another one to coordinate tasks in parallel, maximizing the task execution efficiency. To validate REMAC's effectiveness, we build a multi-agent environment for long-horizon robot manipulation and navigation based on RoboCasa, featuring 4 task categories with 27 task styles and 50+ different objects. Based on it, we further benchmark state-of-the-art reasoning models, including DeepSeek-R1, o3-mini, QwQ, and Grok3, demonstrating REMAC's superiority by boosting average success rates by 40% and execution efficiency by 52.7% over the single robot baseline.
This technical report extends the SIGMOD 2025 paper "A Modular Graph-Native Query Optimization Framework" by providing a comprehensive exposition of GOpt's advanced technical mechanisms, implementation strategies, and extended evaluations. While the original paper introduced GOpt's unified intermediate representation (GIR) and demonstrated its performance benefits, this report delves into the framework's implementation depth: (1) the full specification of GOpt's optimization rules; (2) a systematic treatment of semantic variations (e.g., homomorphism vs. edge-distinct matching) across query languages and their implications for optimization; (3) the design of GOpt's Physical integration interface, enabling seamless integration with transactional (Neo4j) and distributed (GraphScope) backends via engine-specific operator customization; and (4) a detailed analysis of plan transformations for LDBC benchmark queries.
To achieve digital intelligence transformation and carbon neutrality, effective production planning is crucial for integrated refinery-petrochemical complexes. Modern refinery planning relies on advanced optimization techniques, whose development requires reproducible benchmark problems. However, existing benchmarks lack practical context or impose oversimplified assumptions, limiting their applicability to enterprise-wide optimization. To bridge the substantial gap between theoretical research and industrial applications, this paper introduces the first open-source, demand-driven benchmark for industrial-scale refinery-petrochemical complexes with transparent model formulations and comprehensive input parameters. The benchmark incorporates a novel port-stream hybrid superstructure for modular modeling and broad generalizability. Key secondary processing units are represented using the delta-base approach grounded in historical data. Three real-world cases have been constructed to encompass distinct scenario characteristics, respectively addressing (1) a stand-alone refinery without integer variables, (2) chemical site integration with inventory-related integer variables, and (3) multi-period planning. All model parameters are fully accessible. Additionally, this paper provides an analysis of computational performance, ablation experiments on delta-base modeling, and application scenarios for the proposed benchmark.
Robots rely on motion planning to navigate safely and efficiently while performing various tasks. In this paper, we investigate motion planning through Bayesian inference, where motion plans are inferred based on planning objectives and constraints. However, existing Bayesian motion planning methods often struggle to explore low-probability regions of the planning space, where high-quality plans may reside. To address this limitation, we propose the use of heavy-tailed distributions -- specifically, Student's-$t$ distributions -- to enhance probabilistic inferential search for motion plans. We develop a novel sequential single-pass smoothing approach that integrates Student's-$t$ distribution with Monte Carlo sampling. A special case of this approach is ensemble Kalman smoothing, which depends on short-tailed Gaussian distributions. We validate the proposed approach through simulations in autonomous vehicle motion planning, demonstrating its superior performance in planning, sampling efficiency, and constraint satisfaction compared to ensemble Kalman smoothing. While focused on motion planning, this work points to the broader potential of heavy-tailed distributions in enhancing probabilistic decision-making in robotics.
Vision-language-action models (VLAs) have shown potential in leveraging pretrained vision-language models and diverse robot demonstrations for learning generalizable sensorimotor control. While this paradigm effectively utilizes large-scale data from both robotic and non-robotic sources, current VLAs primarily focus on direct input--output mappings, lacking the intermediate reasoning steps crucial for complex manipulation tasks. As a result, existing VLAs lack temporal planning or reasoning capabilities. In this paper, we introduce a method that incorporates explicit visual chain-of-thought (CoT) reasoning into vision-language-action models (VLAs) by predicting future image frames autoregressively as visual goals before generating a short action sequence to achieve these goals. We introduce CoT-VLA, a state-of-the-art 7B VLA that can understand and generate visual and action tokens. Our experimental results demonstrate that CoT-VLA achieves strong performance, outperforming the state-of-the-art VLA model by 17% in real-world manipulation tasks and 6% in simulation benchmarks. Project website: https://cot-vla.github.io/
Ensuring safety in reinforcement learning (RL)-based robotic systems is a critical challenge, especially in contact-rich tasks within unstructured environments. While the state-of-the-art safe RL approaches mitigate risks through safe exploration or high-level recovery mechanisms, they often overlook low-level execution safety, where reflexive responses to potential hazards are crucial. Similarly, variable impedance control (VIC) enhances safety by adjusting the robot's mechanical response, yet lacks a systematic way to adapt parameters, such as stiffness and damping throughout the task. In this paper, we propose Bresa, a Bio-inspired Reflexive Hierarchical Safe RL method inspired by biological reflexes. Our method decouples task learning from safety learning, incorporating a safety critic network that evaluates action risks and operates at a higher frequency than the task solver. Unlike existing recovery-based methods, our safety critic functions at a low-level control layer, allowing real-time intervention when unsafe conditions arise. The task-solving RL policy, running at a lower frequency, focuses on high-level planning (decision-making), while the safety critic ensures instantaneous safety corrections. We validate Bresa on multiple tasks including a contact-rich robotic task, demonstrating its reflexive ability to enhance safety, and adaptability in unforeseen dynamic environments. Our results show that Bresa outperforms the baseline, providing a robust and reflexive safety mechanism that bridges the gap between high-level planning and low-level execution. Real-world experiments and supplementary material are available at project website https://jack-sherman01.github.io/Bresa.
The paper examines social media content to measure and model risk behavior in natural emergencies from an appraisal theory perspective. We calculate individual risk behavior quotients and relate them to individual and peer emotional and actionable cognitive responses for 774 individual Twitter users affected by the Sandy hurricane landfall. We employ vector analysis to compute risk behavior quotients. By utilizing geographic information associated with the tweets, both implicitly and explicitly, we track each user's path and determine the average vector of their movement. The risk quotient is obtained by comparing risk exposure at the origin and destination of the average vector. We assess risk exposure for each zone in the study area by combining pre-hurricane evacuation plans with post-event flooding data, as reported by the National Weather Service. By using the emotional and actionable content of the tweets as predictors for risk, we found that sharing actionable information relates to slightly higher risk exposure. At the same time, overall, the subjects tended to move away from the riskiest areas of the storm. Finally, individuals surrounded by more peers are less likely to be affected, while those surrounded by more tweeting activity are more likely to be affected risk-prone.
Neurosymbolic programs combine deep learning with symbolic reasoning to achieve better data efficiency, interpretability, and generalizability compared to standalone deep learning approaches. However, existing neurosymbolic learning frameworks implement an uneasy marriage between a highly scalable, GPU-accelerated neural component with a slower symbolic component that runs on CPUs. We propose Lobster, a unified framework for harnessing GPUs in an end-to-end manner for neurosymbolic learning. Lobster maps a general neurosymbolic language based on Datalog to the GPU programming paradigm. This mapping is implemented via compilation to a new intermediate language called APM. The extra abstraction provided by APM allows Lobster to be both flexible, supporting discrete, probabilistic, and differentiable modes of reasoning on GPU hardware with a library of provenance semirings, and performant, implementing new optimization passes. We demonstrate that Lobster programs can solve interesting problems spanning the domains of natural language processing, image processing, program reasoning, bioinformatics, and planning. On a suite of 8 applications, Lobster achieves an average speedup of 5.3x over Scallop, a state-of-the-art neurosymbolic framework, and enables scaling of neurosymbolic solutions to previously infeasible tasks.
The staging scenario for CLIC has been updated following new studies of the beam emittance through the accelerator chain, which has resulted in higher expected luminosities, and a change in baseline to a 100 Hz repetition rate at the initial energy stage. Here, the Higgs coupling sensitivities are updated for the new staging plan.
Recent advances in deep thinking models have demonstrated remarkable reasoning capabilities on mathematical and coding tasks. However, their effectiveness in embodied domains which require continuous interaction with environments through image action interleaved trajectories remains largely -unexplored. We present Embodied Reasoner, a model that extends o1 style reasoning to interactive embodied search tasks. Unlike mathematical reasoning that relies primarily on logical deduction, embodied scenarios demand spatial understanding, temporal reasoning, and ongoing self-reflection based on interaction history. To address these challenges, we synthesize 9.3k coherent Observation-Thought-Action trajectories containing 64k interactive images and 90k diverse thinking processes (analysis, spatial reasoning, reflection, planning, and verification). We develop a three-stage training pipeline that progressively enhances the model's capabilities through imitation learning, self-exploration via rejection sampling, and self-correction through reflection tuning. The evaluation shows that our model significantly outperforms those advanced visual reasoning models, e.g., it exceeds OpenAI o1, o3-mini, and Claude-3.7 by +9\%, 24\%, and +13\%. Analysis reveals our model exhibits fewer repeated searches and logical inconsistencies, with particular advantages in complex long-horizon tasks. Real-world environments also show our superiority while exhibiting fewer repeated searches and logical inconsistency cases.
Several hierarchical reinforcement learning methods leverage planning to create a graph or sequences of intermediate goals, guiding a lower-level goal-conditioned (GC) policy to reach some final goals. The low-level policy is typically conditioned on the current goal, with the aim of reaching it as quickly as possible. However, this approach can fail when an intermediate goal can be reached in multiple ways, some of which may make it impossible to continue toward subsequent goals. To address this issue, we introduce two instances of Markov Decision Process (MDP) where the optimization objective favors policies that not only reach the current goal but also subsequent ones. In the first, the agent is conditioned on both the current and final goals, while in the second, it is conditioned on the next two goals in the sequence. We conduct a series of experiments on navigation and pole-balancing tasks in which sequences of intermediate goals are given. By evaluating policies trained with TD3+HER on both the standard GC-MDP and our proposed MDPs, we show that, in most cases, conditioning on the next two goals improves stability and sample efficiency over other approaches.
One of the core components of our world models is 'intuitive physics' - an understanding of objects, space, and causality. This capability enables us to predict events, plan action and navigate environments, all of which rely on a composite sense of objecthood. Despite its importance, there is no single, unified account of objecthood, though multiple theoretical frameworks provide insights. In the first part of this paper, we present a comprehensive overview of the main theoretical frameworks in objecthood research - Gestalt psychology, enactive cognition, and developmental psychology - and identify the core capabilities each framework attributes to object understanding, as well as what functional roles they play in shaping world models in biological agents. Given the foundational role of objecthood in world modelling, understanding objecthood is also essential in AI. In the second part of the paper, we evaluate how current AI paradigms approach and test objecthood capabilities compared to those in cognitive science. We define an AI paradigm as a combination of how objecthood is conceptualised, the methods used for studying objecthood, the data utilised, and the evaluation techniques. We find that, whilst benchmarks can detect that AI systems model isolated aspects of objecthood, the benchmarks cannot detect when AI systems lack functional integration across these capabilities, not solving the objecthood challenge fully. Finally, we explore novel evaluation approaches that align with the integrated vision of objecthood outlined in this paper. These methods are promising candidates for advancing from isolated object capabilities toward general-purpose AI with genuine object understanding in real-world contexts.
Recent advancements in Text-to-SQL, driven by large language models, are democratizing data access. Despite these advancements, enterprise deployments remain challenging due to the need to capture business-specific knowledge, handle complex queries, and meet expectations of continuous improvements. To address these issues, we designed and implemented GenEdit: our Text-to-SQL generation system that improves with user feedback. GenEdit builds and maintains a company-specific knowledge set, employs a pipeline of operators decomposing SQL generation, and uses feedback to update its knowledge set to improve future SQL generations. We describe GenEdit's architecture made of two core modules: (i) decomposed SQL generation; and (ii) knowledge set edits based on user feedback. For generation, GenEdit leverages compounding operators to improve knowledge retrieval and to create a plan as chain-of-thought steps that guides generation. GenEdit first retrieves relevant examples in an initial retrieval stage where original SQL queries are decomposed into sub-statements, clauses or sub-queries. It then also retrieves instructions and schema elements. Using the retrieved contextual information, GenEdit then generates step-by-step plan in natural language on how to produce the query. Finally, GenEdit uses the plan to generate SQL, minimizing the need for model reasoning, which enhances complex SQL generation. If necessary, GenEdit regenerates the query based on syntactic and semantic errors. The knowledge set edits are recommended through an interactive copilot, allowing users to iterate on their feedback and to regenerate SQL queries as needed. Each generation uses staged edits which update the generation prompt. Once the feedback is submitted, it gets merged after passing regression testing and obtaining an approval, improving future generations.
Heavy-flavour physics is an essential component of the particle-physics programme, offering critical tests of the Standard Model and far-reaching sensitivity to physics beyond it. Experiments such as LHCb, Belle II, and BESIII drive progress in the field, along with contributions from ATLAS and CMS. The LHCb Upgrade II and upgraded Belle II experiments will provide unique and highly sensitive measurements for decades, playing a key role in the searches for new physics. Future facilities with significant heavy-flavour capabilities will further expand these opportunities. We advocate for a European Strategy that fully supports Upgrade II of LHCb and an upgrade of Belle II, along with their subsequent exploitation. Additionally, we support a long-term plan that fully integrates flavour physics in an $e^+e^-$ collider to run as a $Z$ factory.
In this paper, we develop a fast mixed-integer convex programming (MICP) framework for multi-robot navigation by combining graph attention networks and distributed optimization. We formulate a mixed-integer optimization problem for receding horizon motion planning of a multi-robot system, taking into account the surrounding obstacles. To address the resulting multi-agent MICP problem in real time, we propose a framework that utilizes heterogeneous graph attention networks to learn the latent mapping from problem parameters to optimal binary solutions. Furthermore, we apply a distributed proximal alternating direction method of multipliers algorithm for solving the convex continuous optimization problem. We demonstrate the effectiveness of our proposed framework through experiments conducted on a robotic testbed.
Existing benchmarks for Vision-Language Model (VLM) on autonomous driving (AD) primarily assess interpretability through open-form visual question answering (QA) within coarse-grained tasks, which remain insufficient to assess capabilities in complex driving scenarios. To this end, we introduce $\textbf{VLADBench}$, a challenging and fine-grained dataset featuring close-form QAs that progress from static foundational knowledge and elements to advanced reasoning for dynamic on-road situations. The elaborate $\textbf{VLADBench}$ spans 5 key domains: Traffic Knowledge Understanding, General Element Recognition, Traffic Graph Generation, Target Attribute Comprehension, and Ego Decision-Making and Planning. These domains are further broken down into 11 secondary aspects and 29 tertiary tasks for a granular evaluation. A thorough assessment of general and domain-specific (DS) VLMs on this benchmark reveals both their strengths and critical limitations in AD contexts. To further exploit the cognitive and reasoning interactions among the 5 domains for AD understanding, we start from a small-scale VLM and train the DS models on individual domain datasets (collected from 1.4M DS QAs across public sources). The experimental results demonstrate that the proposed benchmark provides a crucial step toward a more comprehensive assessment of VLMs in AD, paving the way for the development of more cognitively sophisticated and reasoning-capable AD systems.
Software as a Service (SaaS) pricing models, encompassing features, usage limits, plans, and add-ons, have grown exponentially in complexity, evolving from offering tens to thousands of configuration options. This rapid expansion poses significant challenges for the development and operation of SaaS-based Information Systems (IS), as manual management of such configurations becomes time-consuming, error-prone, and ultimately unsustainable. The emerging paradigm of Pricing-driven DevOps aims to address these issues by automating pricing management tasks, such as transforming human-oriented pricings into machine-oriented (iPricing) or finding the optimal subscription that matches the requirements of a certain user, ultimately reducing human intervention. This paper advances the field by proposing seven analysis operations that partially or fully support these pricing management tasks, thus serving as a foundation for defining new, more specialized operations. To achieve this, we mapped iPricings into Constraint Satisfaction Optimization Problems (CSOP), an approach successfully used in similar domains, enabling us to implement and apply these operations to uncover latent, yet non-trivial insights from complex pricing models. The proposed approach has been implemented in a reference framework using MiniZinc, and tested with over 150 pricing models, identifying errors in 35 pricings of the benchmark. Results demonstrate its effectiveness in identifying errors and its potential to streamline Pricing-driven DevOps.