cs
cs
11-26 00:00
arXiv:2511.19438v1 Announce Type: new Abstract: The increasing adoption of large language model (LLMs) on heterogeneous computing platforms poses significant challenges for achieving high inference efficiency. To address the low inference efficiency of LLMs across diverse heterogeneous platforms, this paper proposes a practical optimization method, Opt4GPTQ, designed for 4-bit GPTQ quantized LLMs inference on heterogeneous AI accelerators. Built upon the vLLM serving system, Opt4GPTQ integrates three platform-level optimization strategies: Shared Memory Buffering optimization (SMB-Opt), which caches data in shared memory and employs single-threaded writes; Vectorized Memory Loading optimization (VML-Opt), which utilizes vectorized memory operations for efficient data loading; and Inline Assembly optimization (ILAOpt), which directly leverages hardware-native vector halfprecision addition and fused multiply-accumulate instructions for efficient execution. Experimental results show that Opt4GPTQ effectively improves inference performance across different models, achieving up to 84.42% throughput improvement and up to 51.35% latency reduction. This work highlights the critical role of platform-level engineering optimizations in enabling efficient LLMs inference on emerging heterogeneous AI acceleration architectures and provides valuable deployment experience and methodologies for future heterogeneous platform adaptation.
cs.dccs.pf
cs
cs
11-26 00:00
arXiv:2511.19445v1 Announce Type: new Abstract: We propose a parallel shared-memory schema to cooperatively optimize the solution of a Capacitated Vehicle Routing Problem instance with minimal synchronization effort and without the need for an explicit decomposition. To this end, we design FILO2$^x$ as a single-trajectory parallel adaptation of the FILO2 algorithm originally proposed for extremely large-scale instances and described in Accorsi and Vigo (2024). Using the locality of the FILO2 optimization applications, in FILO2$^x$ several possibly unrelated solution areas are concurrently asynchronously optimized. The overall search trajectory emerges as an iteration-based parallelism obtained by the simultaneous optimization of the same underlying solution performed by several solvers. Despite the high efficiency exhibited by the single-threaded FILO2 algorithm, the computational results show that, by better exploiting the available computing resources, FILO2$^x$ can greatly enhance the resolution time compared to the original approach, still maintaining a similar final solution quality for instances ranging from hundreds to hundreds of thousands customers.
cs.dccs.dm
cs
cs
11-26 00:00
arXiv:2511.19446v1 Announce Type: new Abstract: This paper presents a novel class of information-theoretic strategies for solving the game of Mastermind, achieving state-of-the-art performance among known heuristic methods. The core contribution is the application of a weighted entropy heuristic, based on the Belis-Guias, u framework, which assigns context- dependent utility values to each of the possible feedback types. A genetic algorithm optimization approach discovers interpret-able weight patterns that reflect strategic game dynamics. First, I demonstrate that a single, fixed vector of optimized weights achieves a remarkable 4.3565 average guesses with a maximum of 5. Building upon this, I introduce a stage-weighted heuristic with distinct utility vectors for each turn, achieving 4.3488 average guesses with a maximum of 6, approaching the theoretical optimum of 4.3403 by less than 0.2%. The method retains the computational efficiency of classical one-step-ahead heuristics while significantly improving performance through principled information valuation. A complete implementation and all optimized parameters are provided for full reproducibility.
cs.itcs.gtmath.it
cs
cs
11-26 00:00
arXiv:2511.19449v1 Announce Type: new Abstract: Electrifying passenger cars will impact future power systems. To understand the challenges and opportunities that arise, it is necessary to reflect "sector coupling" in the modeling space. This paper focuses on a specific modeling approach that includes dozens of individual BEV profiles rather than one aggregated BEV profile. Although including additional BEV profiles increases model complexity and runtime, it avoids losing information in the aggregation process. We investigate how many profiles are needed to ensure the accuracy of the results and the extent to which fewer profiles can be traded for runtime efficiency gains. We also examine whether selecting specific profiles influences optimal results. We demonstrate that including too few profiles may result in distorted optimal solutions. However, beyond a certain threshold, adding more profiles does not significantly enhance the robustness of the results. More generally, for fleets of 5 to 20 million BEVs, we derive a rule of thumb consisting in including enough profiles such that each profile represents 200,000 to 250,000 vehicles, ensuring accurate results without excessive runtime.
eess.sycs.sy
cs
cs
11-26 00:00
arXiv:2511.19451v1 Announce Type: new Abstract: The paper addresses a continuous-time continuous-space chance-constrained stochastic optimal control (SOC) problem where the probability of failure to satisfy given state constraints is explicitly bounded. We leverage the notion of exit time from continuous-time stochastic calculus to formulate a chance-constrained SOC problem. Without any conservative approximation, the chance constraint is transformed into an expectation of an indicator function which can be incorporated into the cost function by considering a dual formulation. We then express the dual function in terms of the solution to a Hamilton-Jacobi-Bellman partial differential equation parameterized by the dual variable. Under a certain assumption on the system dynamics and cost function, it is shown that a strong duality holds between the primal chance-constrained problem and its dual. The Path integral approach is utilized to numerically solve the dual problem via gradient ascent using open-loop samples of system trajectories. We present simulation studies on chance-constrained motion planning for spatial navigation of mobile robots and the solution of the path integral approach is compared with that of the finite difference method.
cs.roeess.sycs.sy
cs
cs
11-26 00:00
arXiv:2511.19452v1 Announce Type: new Abstract: This paper presents a closed-loop framework for conflict-free routing and scheduling of multi-aircraft in Terminal Manoeuvring Areas (TMA), aimed at reducing congestion and enhancing landing efficiency. Leveraging data-driven arrival inputs (either historical or predicted), we formulate a mixed-integer optimization model for real-time control, incorporating an extended TMA network spanning a 50-nautical-mile radius around Changi Airport. The model enforces safety separation, speed adjustments, and holding time constraints while maximizing runway throughput. A rolling-horizon Model Predictive Control (MPC) strategy enables closed-loop integration with a traffic simulator, dynamically updating commands based on real-time system states and predictions. Computational efficiency is validated across diverse traffic scenarios, demonstrating a 7-fold reduction in computation time during peak congestion compared to onetime optimization, using Singapore ADS-B dataset. Monte Carlo simulations under travel time disturbances further confirm the framework's robustness. Results highlight the approach's operational resilience and computational scalability, offering actionable decision support for Air Traffic Controller Officers (ATCOs) through real-time optimization and adaptive replanning.
cs.maeess.sycs.sy
cs
cs
11-26 00:00
arXiv:2511.19453v1 Announce Type: new Abstract: Autonomous vehicles (AVs) are evolving into mobile computing platforms, equipped with powerful processors and diverse sensors that generate massive heterogeneous data, for example 14 TB per day. Supporting emerging third-party applications calls for a general-purpose, queryable onboard storage system. Yet today's data loggers and storage stacks in vehicles fail to deliver efficient data storage and retrieval. This paper presents AVS, an Autonomous Vehicle Storage system that co-designs computation with a hierarchical layout: modality-aware reduction and compression, hot-cold tiering with daily archival, and a lightweight metadata layer for indexing. The design is grounded with system-level benchmarks on AV data that cover SSD and HDD filesystems and embedded indexing, and is validated on embedded hardware with real L4 autonomous driving traces. The prototype delivers predictable real-time ingest, fast selective retrieval, and substantial footprint reduction under modest resource budgets. The work also outlines observations and next steps toward more scalable and longer deployments to motivate storage as a first-class component in AV stacks.
cs.rocs.dbcs.dccs.os
cs
cs
11-26 00:00
arXiv:2511.19454v1 Announce Type: new Abstract: The Multi-Traveling Salesman Problem (MTSP) is a commonly used mathematical model for multi-agent task allocation. However, as the number of agents and task targets increases, existing optimization-based methods often incur prohibitive computational costs, posing significant challenges to large-scale coordination in unmanned systems. To address this issue, this paper proposes a K-means-inspired task allocation framework that reformulates the MTSP as a spatially constrained classification process. By leveraging spatial coherence, the proposed method enables fast estimation of path costs and efficient task grouping, thereby fundamentally reducing overall computational complexity. Extensive simulation results demonstrate that the framework can maintain high solution quality even in extremely large-scale scenarios-for instance, in tasks involving 1000 agents and 5000 targets. The findings indicate that this "cluster-then-route" decomposition strategy offers an efficient and reliable solution for large-scale multi-agent task allocation.
cs.roeess.sycs.sy
cs
cs
11-26 00:00
arXiv:2511.19456v1 Announce Type: new Abstract: Complex computational problems in science often consist of smaller parts that can have largely distinct compute requirements from one another. For optimal efficiency, analyzing each subtask and scheduling it on the best-suited hardware would be necessary. Other considerations must be taken into account, too, such as parallelism, dependencies between different subtasks, and data transfer speeds between devices. To achieve this, directed acyclic graphs are often employed to represent these problems and enable utilizing as much hardware as possible on a given machine. In this paper, we present a software framework written in Julia capable of automatically and dynamically producing statically scheduled and compiled code. We lay theoretical foundations and add domain-specific information about the computation to the existing concepts of DAG scheduling, enabling optimizations that would otherwise be impossible. To illustrate the theory we implement an example application: the computation of matrix elements for scattering processes with many external particles in quantum electrodynamics.
cs.dccs.pf
cs
cs
11-26 00:00
arXiv:2511.19457v1 Announce Type: new Abstract: The resource demands of deep neural network (DNN) models introduce significant performance challenges, especially when deployed on resource-constrained edge devices. Existing solutions like model compression often sacrifice accuracy, while specialized hardware remains costly and inflexible. Hybrid inference methods, however, typically overlook how operator characteristics impact performance. In this work, we present SparOA, a CPU-GPU hybrid inference framework, which leverages both sparsity and computational intensity to optimize operator scheduling. SparOA embraces aforementioned challenges through three key components: (1) a threshold predictor that accurately determines optimal sparsity and computational intensity thresholds; (2) a reinforcement learning-based scheduler that dynamically optimizes resource allocation based on real-time hardware states; and (3) a hybrid inference engine that enhances efficiency through asynchronous execution and batch size optimization.Extensive results show that SparOA achieves an average speedup of 1.22-1.31x compared to all baselines, and outperforms the CPU-Only by up to 50.7x. Also, SparOA achieves optimal energy-per-inference, consuming 7\%-16\% less energy than the SOTA co-execution baseline.
cs.dccs.ai
cs
cs
11-26 00:00
arXiv:2511.19458v1 Announce Type: new Abstract: Recent text-to-image (T2I) models generate semantically coherent images from textual prompts, yet evaluating how well they align with individual user preferences remains an open challenge. Conventional evaluation methods, general reward functions or similarity-based metrics, fail to capture the diversity and complexity of personal visual tastes. In this work, we present PIGReward, a personalized reward model that dynamically generates user-conditioned evaluation dimensions and assesses images through CoT reasoning. To address the scarcity of user data, PIGReward adopt a self-bootstrapping strategy that reasons over limited reference data to construct rich user contexts, enabling personalization without user-specific training. Beyond evaluation, PIGReward provides personalized feedback that drives user-specific prompt optimization, improving alignment between generated images and individual intent. We further introduce PIGBench, a per-user preference benchmark capturing diverse visual interpretations of shared prompts. Extensive experiments demonstrate that PIGReward surpasses existing methods in both accuracy and interpretability, establishing a scalable and reasoning-based foundation for personalized T2I evaluation and optimization. Taken together, our findings highlight PIGReward as a robust steptoward individually aligned T2I generation.
cs.cvcs.ai
cs
cs
11-26 00:00
arXiv:2511.19460v1 Announce Type: new Abstract: Smart grid technological advances present a recent class of complex interdisciplinary modeling and increasingly difficult simulation problems to solve using traditional computational methods. To simulate a smart grid requires a systemic approach to integrated modeling of power systems, energy markets, demand-side management, and much other resources and assets that are becoming part of the current paradigm of the power grid. This paper presents a backbone model of a smart grid to test alternative scenarios for the grid. This tool simulates disparate systems to validate assumptions before the human scale model. Thanks to a distributed optimization of subsystems, the production and consumption scheduling is achieved while maintaining flexibility and scalability.
cs.dccs.syeess.sycs.ai
cs
cs
11-26 00:00
arXiv:2511.19463v1 Announce Type: new Abstract: Urban Building Energy Modeling (UBEM) plays a central role in understanding and forecasting energy consumption at the city scale. In this work, we present a UBEM pipeline that integrates EnergyPlus simulations, high-performance computing (HPC), and open geospatial datasets to estimate the energy demand of buildings in Bologna, Italy. Geometric information including building footprints and heights was obtained from the Bologna Open Data portal and enhanced with aerial LiDAR measurements. Non-geometric attributes such as construction materials, insulation characteristics, and window performance were derived from regional building regulations and the European TABULA database. The computation was carried out on Leonardo, the Cineca-hosted supercomputer, enabling the simulation of approximately 25,000 buildings in under 30 minutes.
cs.dcphysics.app-ph
cs
cs
11-26 00:00
arXiv:2511.19464v1 Announce Type: new Abstract: SOCs and CSIRTs face increasing pressure to automate incident categorization, yet the use of cloud-based LLMs introduces costs, latency, and confidentiality risks. We investigate whether locally executed SLMs can meet this challenge. We evaluated 21 models ranging from 1B to 20B parameters, varying the temperature hyperparameter and measuring execution time and precision across two distinct architectures. The results indicate that temperature has little influence on performance, whereas the number of parameters and GPU capacity are decisive factors.
cs.lgcs.dccs.crcs.pfcs.ai
cs
cs
11-26 00:00
arXiv:2511.19465v1 Announce Type: new Abstract: Nowadays, social networks are becoming a popular way of analyzing tourist behavior, thanks to the digital traces left by travelers during their stays on these networks. The massive amount of data generated; by the propensity of tourists to share comments and photos during their trip; makes it possible to model their journeys and analyze their behavior. Predicting the next movement of tourists plays a key role in tourism marketing to understand demand and improve decision support. In this paper, we propose a method to understand and to learn tourists' movements based on social network data analysis to predict future movements. The method relies on a machine learning grammatical inference algorithm. A major contribution in this paper is to adapt the grammatical inference algorithm to the context of big data. Our method produces a hidden Markov model representing the movements of a group of tourists. The hidden Markov model is flexible and editable with new data. The capital city of France, Paris is selected to demonstrate the efficiency of the proposed methodology.
cs.lgcs.ai
cs
cs
11-26 00:00
arXiv:2511.19466v1 Announce Type: new Abstract: Approximating training-point influence on test predictions is critical for deploying deep-learning vision models, essential for locating noisy data. Though the influence function was proposed for attributing how infinitesimal up-weighting or removal of individual training examples affects model outputs, its implementation is still challenging in deep-learning vision models: inverse-curvature computations are expensive, and training non-stationarity invalidates static approximations. Prior works use iterative solvers and low-rank surrogates to reduce cost, but offline computation lags behind training dynamics, and missing confidence calibration yields fragile rankings that misidentify critical examples. To address these challenges, we introduce a Stability-Guided Online Influence Framework (SG-OIF), the first framework that treats algorithmic stability as a real-time controller, which (i) maintains lightweight anchor IHVPs via stochastic Richardson and preconditioned Neumann; (ii) proposes modular curvature backends to modulate per-example influence scores using stability-guided residual thresholds, anomaly gating, and confidence. Experimental results show that SG-OIF achieves SOTA (State-Of-The-Art) on noise-label and out-of-distribution detection tasks across multiple datasets with various corruption. Notably, our approach achieves 91.1\% accuracy in the top 1\% prediction samples on the CIFAR-10 (20\% asym), and gets 99.8\% AUPR score on MNIST, effectively demonstrating that this framework is a practical controller for online influence estimation.
cs.lgcs.cvcs.ai
cs
cs
11-26 00:00
arXiv:2511.19468v1 Announce Type: new Abstract: If AI is a foundational general-purpose technology, we should anticipate that demand for AI compute -- and energy -- will continue to grow. The Sun is by far the largest energy source in our solar system, and thus it warrants consideration how future AI infrastructure could most efficiently tap into that power. This work explores a scalable compute system for machine learning in space, using fleets of satellites equipped with solar arrays, inter-satellite links using free-space optics, and Google tensor processing unit (TPU) accelerator chips. To facilitate high-bandwidth, low-latency inter-satellite communication, the satellites would be flown in close proximity. We illustrate the basic approach to formation flight via a 81-satellite cluster of 1 km radius, and describe an approach for using high-precision ML-based models to control large-scale constellations. Trillium TPUs are radiation tested. They survive a total ionizing dose equivalent to a 5 year mission life without permanent failures, and are characterized for bit-flip errors. Launch costs are a critical part of overall system cost; a learning curve analysis suggests launch to low-Earth orbit (LEO) may reach $\lesssim$\$200/kg by the mid-2030s.
cs.lgcs.dc
cs
cs
11-26 00:00
arXiv:2511.19470v1 Announce Type: new Abstract: Quantifying modality contributions in multimodal models remains a challenge, as existing approaches conflate the notion of contribution itself. Prior work relies on accuracy-based approaches, interpreting performance drops after removing a modality as indicative of its influence. However, such outcome-driven metrics fail to distinguish whether a modality is inherently informative or whether its value arises only through interaction with other modalities. This distinction is particularly important in cross-attention architectures, where modalities influence each other's representations. In this work, we propose a framework based on Partial Information Decomposition (PID) that quantifies modality contributions by decomposing predictive information in internal embeddings into unique, redundant, and synergistic components. To enable scalable, inference-only analysis, we develop an algorithm based on the Iterative Proportional Fitting Procedure (IPFP) that computes layer and dataset-level contributions without retraining. This provides a principled, representation-level view of multimodal behavior, offering clearer and more interpretable insights than outcome-based metrics.
cs.lgcs.aics.cl
cs
cs
11-26 00:00
arXiv:2511.19472v1 Announce Type: new Abstract: Prefix adders are widely used in compute-intensive applications for their high speed. However, designing optimized prefix adders is challenging due to strict design rules and an exponentially large design space. We introduce PrefixGPT, a generative pre-trained Transformer (GPT) that directly generates optimized prefix adders from scratch. Our approach represents an adder's topology as a two-dimensional coordinate sequence and applies a legality mask during generation, ensuring every design is valid by construction. PrefixGPT features a customized decoder-only Transformer architecture. The model is first pre-trained on a corpus of randomly synthesized valid prefix adders to learn design rules and then fine-tuned to navigate the design space for optimized design quality. Compared with existing works, PrefixGPT not only finds a new optimal design with a 7.7% improved area-delay product (ADP) but exhibits superior exploration quality, lowering the average ADP by up to 79.1%. This demonstrates the potential of GPT-style models to first master complex hardware design principles and then apply them for more efficient design optimization.
cs.lgcs.arcs.ai
cs
cs
11-26 00:00
arXiv:2511.19473v1 Announce Type: new Abstract: Diffusion Language Models (DLMs) have shown strong potential for text generation and are becoming a competitive alternative to autoregressive models. The denoising strategy plays an important role in determining the quality of their outputs. Mainstream denoising strategies include Standard Diffusion and BlockDiffusion. Standard Diffusion performs global denoising without restricting the update range, often finalizing incomplete context and causing premature end-of-sequence predictions. BlockDiffusion updates fixed-size blocks in a preset order, but its rigid structure can break apart coherent semantic units and disrupt reasoning. We present WavefrontDiffusion, a dynamic decoding approach that expands a wavefront of active tokens outward from finalized positions. This adaptive process follows the natural flow of semantic structure while keeping computational cost equal to block-based methods. Across four benchmarks in reasoning and code generation, WavefrontDiffusion achieves state-of-the-art performance while producing outputs with higher semantic fidelity, showing the value of adaptive scheduling for more coherent and efficient generation.
cs.lgcs.ai
cs
cs
11-26 00:00
arXiv:2511.19474v1 Announce Type: new Abstract: Automatically detecting abnormal events in videos is crucial for modern autonomous systems, yet existing Video Anomaly Detection (VAD) benchmarks lack the scene diversity, balanced anomaly coverage, and temporal complexity needed to reliably assess real-world performance. Meanwhile, the community is increasingly moving toward Video Anomaly Understanding (VAU), which requires deeper semantic and causal reasoning but remains difficult to benchmark due to the heavy manual annotation effort it demands. In this paper, we introduce Pistachio, a new VAD/VAU benchmark constructed entirely through a controlled, generation-based pipeline. By leveraging recent advances in video generation models, Pistachio provides precise control over scenes, anomaly types, and temporal narratives, effectively eliminating the biases and limitations of Internet-collected datasets. Our pipeline integrates scene-conditioned anomaly assignment, multi-step storyline generation, and a temporally consistent long-form synthesis strategy that produces coherent 41-second videos with minimal human intervention. Extensive experiments demonstrate the scale, diversity, and complexity of Pistachio, revealing new challenges for existing methods and motivating future research on dynamic and multi-event anomaly understanding.
cs.mmcs.cvcs.ai
cs
cs
11-26 00:00
arXiv:2511.19475v1 Announce Type: new Abstract: Tracking and segmentation play essential roles in video understanding, providing basic positional information and temporal association of objects within video sequences. Despite their shared objective, existing approaches often tackle these tasks using specialized architectures or modality-specific parameters, limiting their generalization and scalability. Recent efforts have attempted to unify multiple tracking and segmentation subtasks from the perspectives of any modality input or multi-task inference. However, these approaches tend to overlook two critical challenges: the distributional gap across different modalities and the feature representation gap across tasks. These issues hinder effective cross-task and cross-modal knowledge sharing, ultimately constraining the development of a true generalist model. To address these limitations, we propose a universal tracking and segmentation framework named SATA, which unifies a broad spectrum of tracking and segmentation subtasks with any modality input. Specifically, a Decoupled Mixture-of-Expert (DeMoE) mechanism is presented to decouple the unified representation learning task into the modeling process of cross-modal shared knowledge and specific information, thus enabling the model to maintain flexibility while enhancing generalization. Additionally, we introduce a Task-aware Multi-object Tracking (TaMOT) pipeline to unify all the task outputs as a unified set of instances with calibrated ID information, thereby alleviating the degradation of task-specific knowledge during multi-task training. SATA demonstrates superior performance on 18 challenging tracking and segmentation benchmarks, offering a novel perspective for more generalizable video understanding.
cs.mmcs.cvcs.ai
cs
cs
11-26 00:00
arXiv:2511.19448v1 Announce Type: new Abstract: Reliable perception of the environment is a key enabler for autonomous systems, where calibration and localization tasks often rely on robust visual markers. We introduce the PuzzlePole, a new type of fiducial markers derived from the recently proposed PuzzleBoard calibration pattern. The PuzzlePole is a cylindrical marker, enabling reliable recognition and pose estimation from 360{\deg} viewing direction. By leveraging the unique combinatorial structure of the PuzzleBoard pattern, PuzzlePoles provide a high accuracy in localization and orientation while being robust to occlusions. The design offers flexibility for deployment in diverse autonomous systems scenarios, ranging from robot navigation and SLAM to tangible interfaces.
cs.cv
cs
cs
11-26 00:00
arXiv:2511.19450v1 Announce Type: new Abstract: Sharding has emerged as a key technique to address blockchain scalability by partitioning the ledger into multiple shards that process transactions in parallel. Although this approach improves throughput, static or heuristic shard allocation often leads to workload skew, congestion, and excessive cross-shard communication diminishing the scalability benefits of sharding. To overcome these challenges, we propose the Predictive Shard Allocation Protocol (PSAP), a dynamic and intelligent allocation framework that proactively assigns accounts and transactions to shards based on workload forecasts. PSAP integrates a Temporal Workload Forecasting (TWF) model with a safety-constrained reinforcement learning (Safe-PPO) controller, jointly enabling multi-block-ahead prediction and adaptive shard reconfiguration. The protocol enforces deterministic inference across validators through a synchronized quantized runtime and a safety gate that limits stake concentration, migration gas, and utilization thresholds. By anticipating hotspot formation and executing bounded, atomic migrations, PSAP achieves stable load balance while preserving Byzantine safety. Experimental evaluation on heterogeneous datasets, including Ethereum, NEAR, and Hyperledger Fabric mapped via address-clustering heuristics, demonstrates up to 2x throughput improvement, 35\% lower latency, and 20\% reduced cross-shard overhead compared to existing dynamic sharding baselines. These results confirm that predictive, deterministic, and security-aware shard allocation is a promising direction for next-generation scalable blockchain systems.
cs.dc