arXiv subjects

Arvind Easwaran

Publications and source records attributed to Arvind Easwaran.

At least 19 recordsLinked to original sources

arXiv2026

Building real-time digital twin instances with Function+Data Flow: user evaluation and extension for iterative pipelines

Digital twins (DTs) increasingly leverage artificial intelligence (AI) and machine learning (ML) pipelines, both to build real-time DTs from high-fidelity simulations and to instantiate them with historical data. However, engineering these pipelines remains largely ad-hoc: pipelines are hard to specify, validate, and reuse, with scarce dedicated tooling. Function+Data Flow (FDF) addresses this by defining a visual domain-specific language (DSL) that represents functions (ML models) explicitly, enabling their composition and reuse. We implemented FDF in DesCartes Builder, an integrated modeling environment supporting FDF-based DT synthesis and validation. In this paper, we report on an empirical user study evaluating whether FDF and DesCartes Builder can make AI-based DT development more accessible and reliable. Participants implemented a representative real-time DT prototype within DesCartes Builder, and we measured perceived usability and feature adequacy through quantitative and qualitative measures. Our results indicate that DesCartes Builder and FDF achieve a good level of usability across a broad range of potential users, and particularly for the intended audience of domain experts. The study additionally surfaces concrete strengths and areas for improvement of both the tool and the underlying FDF framework. Informed by these findings, we propose H-FDF, a Hierarchical extension of FDF supporting iterative and modular pipelines, enabling the formal specification of more complex DT pipelines such as dual training. Our findings suggest that integrated, model-driven platforms are a promising direction to transform AI-based DT engineering into a disciplined modeling practice.

cs.SE↗

arXiv2026

HyperCut: Fast Inter-Layer Scheduling via Directed Hypergraph and Early Filtering

As deep neural networks (DNNs) continue to scale, inter-layer scheduling, which orchestrates the spatial allocation of compute resources and the temporal execution order across layers, has become a decisive factor in sustaining high utilization and energy efficiency on tiled accelerators. However, existing inter-layer schedulers defer cost feedback until a complete fine-grained intra-layer scheduling has been resolved. The resulting decoupled flow repeatedly explores sub-optimal or even infeasible inter-layer schedules, and the absence of early pruning during the inter-layer phase remains a critical bottleneck for design-space exploration (DSE) in DNN compilers. Our key observation is that the cost of an intra-layer scheduling can be tightly upper-bounded once the inter-layer cut fixes the sub-mesh shape, which lets us cost every inter-layer candidate without solving the intra-layer problem. Hence, we propose a hierarchical partitioning-and-mapping framework, HyperCut, that enables early filtering of inter-layer schedules based on hypergraph partitioning. Based on the directed hypergraph (DHG) abstraction of DNN, we introduce a unified representation, State, that jointly encodes the DHG partition, tile mesh allocation and tensor batch splitting. Thereby, partitioning and mapping are coupled into a union optimization object. For a DNN with N layers, the resulting theoretical design space is bounded by O(N), compared with O(9.899^N) for the state-of-the-art open-source scheduler SET. Across 10 evaluated cases, HyperCut achieves 2.0x performance improvement and 80.47% exploration time reduction over the SET baseline, measured by geometric mean.

cs.AR↗

arXiv2026

Neurosymbolic Reasoning with Incremental Knowledge for Sample Efficient Hierarchical Reinforcement Learning

(Flat) Reinforcement Learning (RL) agents face significant challenges in environments with sparse rewards that require long-horizon reasoning. A compelling approach to improve sample efficiency is to incorporate knowledge into learning and decision-making. In standard Hierarchical RL (HRL), knowledge is encoded in a fixed, non-updatable form, such as architectural choices, and remains unchanged throughout learning. With fixed HRL, reasoning with incremental knowledge learned during exploration is impractical before sufficient environmental knowledge is acquired, leading to poor sample efficiency. In this work, we propose neurosymbolic HRL with {\em Incremental Knowledge (InK)}: symbolic high-level components perform {\em symbolic planning} (e.g. using $D^*$) on an updatable representation of current InK, while low-level goal-conditioned neural modules learn motion primitives through experience using reward shaping. Experiments on navigation tasks demonstrate that incorporating InK substantially improves sample efficiency. Additionally, to perform {\em optimal} symbolic planning given {\em prior} knowledge about the world, we develop Belief World Tree Search. The code is available at https://github.com/CPS-research-group/ink_bwts.

cs.AI↗

arXiv2026

A Hierarchical Stochastic Model Predictive Control Framework for Integrated Request-aware Charge Scheduling and Service Allocation

This study introduces a unified control mechanism for integrated charging and service allocation of Mobility-on-Demand Electric Vehicles (MoD-EVs) operating under a flexible A-to-B rental model. The charging operation of MoD-EVs is influenced by stochastic customer rental requests and time-varying electricity prices. The framework captures the nonlinear battery dynamics and optimizes the charging operations against dynamic electricity prices and to meet stochastic customer demand while prioritizing battery health and cost efficiency. To address the computational challenges posed by the nonlinear battery dynamics, the framework employs a piece-wise linear model integrated into a multi-objective, chance-constrained mixed-integer linear programming (MILP) Model Predictive Control (MPC) formulation. A mixed logical switching mechanism is utilized to determine optimal charging sequences. Furthermore, a distributed approach is implemented to ensure computational scalability compared to centralized alternatives. Evaluation of this approach using a state-of-the-art commercial solver with stochastic EV rental requests under different confidence levels and time-varying electricity prices demonstrates significant benefits of the integrated mechanism design, including a reduction in charging costs and battery capacity degradation compared to the prevailing business-as-usual (BAU) approach and state of the art Laxity-based charging (LC) approach.

eess.SY↗

arXiv2026

SCoReT: Super-Resolution Compression and Reconstruction of Turbulent Flows

High-fidelity simulations of the Navier--Stokes equations (NSE) generate massive amounts of data, motivating the need for efficient compression and reconstruction strategies for turbulent flows. At the same time, reconstructing flow fields from sparse measurements while retaining spectral content, turbulence statistics, and coherent structures remains a major challenge. This work investigates two complementary paradigms for turbulent flow reconstruction: supervised reconstruction and physics-informed reconstruction, in the context of transition to turbulence induced by three-dimensional distributed roughness elements. We introduce a vorticity-augmented supervised approach and a physics-informed approach, implemented through a partially assisted compressible PINN formulation based on the three-dimensional unsteady compressible Navier--Stokes equations. Reconstruction performance is evaluated at different sparsity levels using instantaneous velocity fields, mean-squared error, energy spectra, Reynolds stresses, turbulent kinetic energy, and Q-criterion isosurfaces. Rather than establishing a universal winner, the present study characterises the respective strengths, limitations, and operating regimes of these two approaches. The results indicate that at lower sparsity levels, the vorticity-augmented supervised model yields the lowest reconstruction error, recovers key statistical and spectral features, and enables substantial data compression. The PINN shows potential to reconstruct turbulent flows from sparse measurements without high-resolution labels and exhibits comparatively stable held-out extrapolation behaviour at higher sparsity. These results suggest the potential of combining data-driven and physics-informed learning for flow data compression and physics-informed reconstruction of turbulent flows from sparse data.

physics.flu-dyn↗

arXiv2026

Scenario Generation for Risk-Aware Reinforcement Learning with Probably Approximately Safe Guarantees

Guaranteeing safety is critical to the deployment of reinforcement learning (RL) agents in the real-world, especially as policies learned using deep RL may demonstrate susceptibility to transition perturbations that result in unknown or unsafe behaviour. A method of policy verification is to construct probabilistic barrier-certificates by sampling policy trajectories with respect to safety constraints, thereby demarcating known safe behaviour from unknown behaviour. Obtaining tight upper and lower bounds on the probability of violation of these constraints may be difficult if the policy is susceptible to transition uncertainty or perturbation that places the agent in insufficiently explored states. To address this, we approximate the distribution of the encountered state-space using a variational autoencoder (VAE) and construct upper and lower-bound barrier-certificates using latent characteristics of states to optimize for regions of known, safe behaviour with high confidence. We frame this in our work as a dual optimization problem where the lower-bound barrier-certificate presents a more conservative estimate of the safe region than the upper-bound barrier-certificate. Sampling states that lie within the set difference of the two during training, i.e. the non-robust region, allows us to tighten the upper and lower bounds to provide sharper probabilistic guarantees on safety. Within our study, we describe the guarantees placed and demonstrate the tightness of our bounds experimentally.

cs.LG↗

arXiv2026

Approximation-Free Differentiable Oblique Decision Trees

Decision Trees (DTs) are widely used in safety-critical domains such as medical diagnosis, valued for their interpretability and effectiveness on tabular data. However, training accurate oblique DTs is challenging due to complex optimization landscapes and overfitting risks, particularly in regression. Recent advances have introduced differentiable formulations that enable gradient-based training and joint optimization of decision boundaries and leaf regressors. Yet, existing approaches typically rely on approximations, either through probabilistic softening of boundaries (soft DTs) or quantized gradients such as the Straight-Through Estimator (STE). To overcome these limitations, we propose DTSemNet, a novel, semantically equivalent, and invertible representation of hard oblique DTs as neural networks. DTSemNet enables end-to-end training with standard gradient descent, eliminating the need for approximations in both classification and regression. While classification aligns naturally with this formulation, regression remains challenging due to the joint optimization of internal nodes and leaf regressors. To address this, we analyze the limitations of STE and introduce an annealed Top-k method that provides accurate gradient signals without approximation. Extensive experiments on classification and regression benchmarks show that DTSemNet-trained oblique DTs outperform state-of-the-art differentiable DTs. Furthermore, we demonstrate that DTSemNet can serve as programmatic DT policies in reinforcement learning environments, thereby broadening their applicability.

cs.LG↗

arXiv2025

Real-Time Service Subscription and Adaptive Offloading Control in Vehicular Edge Computing

Vehicular Edge Computing (VEC) has emerged as a promising paradigm for enhancing the computational efficiency and service quality in intelligent transportation systems by enabling vehicles to wirelessly offload computation-intensive tasks to nearby Roadside Units. However, efficient task offloading and resource allocation for time-critical applications in VEC remain challenging due to constrained network bandwidth and computational resources, stringent task deadlines, and rapidly changing network conditions. To address these challenges, we formulate a Deadline-Constrained Task Offloading and Resource Allocation Problem (DOAP), denoted as $\mathbf{P}$, in VEC with both bandwidth and computational resource constraints, aiming to maximize the total vehicle utility. To solve $\mathbf{P}$, we propose $\mathtt{SARound}$, an approximation algorithm based on Linear Program rounding and local-ratio techniques, that improves the best-known approximation ratio for DOAP from $\frac{1}{6}$ to $\frac{1}{4}$. Additionally, we design an online service subscription and offloading control framework to address the challenges of short task deadlines and rapidly changing wireless network conditions. To validate our approach, we develop a comprehensive VEC simulator, VecSim, using the open-source simulation libraries OMNeT++ and Simu5G. VecSim integrates our designed framework to manage the full life-cycle of real-time vehicular tasks. Experimental results, based on profiled object detection applications and real-world taxi trace data, show that $\mathtt{SARound}$ consistently outperforms state-of-the-art baselines under varying network conditions while maintaining runtime efficiency.

cs.DC↗

arXiv2025

Disentangled and Distilled Encoder for Out-of-Distribution Reasoning with Rademacher Guarantees

Recently, the disentangled latent space of a variational autoencoder (VAE) has been used to reason about multi-label out-of-distribution (OOD) test samples that are derived from different distributions than training samples. Disentangled latent space means having one-to-many maps between latent dimensions and generative factors or important characteristics of an image. This paper proposes a disentangled distilled encoder (DDE) framework to decrease the OOD reasoner size for deployment on resource-constrained devices while preserving disentanglement. DDE formalizes student-teacher distillation for model compression as a constrained optimization problem while preserving disentanglement with disentanglement constraints. Theoretical guarantees for disentanglement during distillation based on Rademacher complexity are established. The approach is evaluated empirically by deploying the compressed model on an NVIDIA

cs.LG↗

arXiv2025

PAC One-Step Safety Certification for Black-Box Discrete-Time Stochastic Systems

This paper investigates the problem of safety certification for black-box discrete-time stochastic systems, where both the system dynamics and disturbance distributions are unknown, and only sampled data are available. Under such limited information, ensuring robust or classical quantitative safety over finite or infinite horizons is generally infeasible. To address this challenge, we propose a data-driven framework that provides theoretical one-step safety guarantees in the Probably Approximately Correct (PAC) sense. This one-step guarantee can be applied recursively at each time step, thereby yielding step-by-step safety assurances over extended horizons. Our approach formulates barrier certificate conditions based solely on sampled data and establishes PAC safety guarantees by leveraging the VC dimension, scenario approaches, Markov's inequality, and Hoeffding's inequality. Two sampling procedures are proposed, and three methods are proposed to derive PAC safety guarantees. The properties and comparative advantages of these three methods are thoroughly discussed. Finally, the effectiveness of the proposed methods are demonstrated through several numerical examples.

eess.SY↗

arXiv2025

Managing Charging Induced Grid Stress and Battery Degradation in Electric Taxi Fleets

Operating fleets of electric vehicles (EVs) introduces several challenges, some of which are borne by the fleet operator, and some of which are borne by the power grid. To maximize short-term profit a fleet operator could always charge EVs at the maximum rate to ensure vehicles are ready to service ride demand. However, due to the stochastic nature of electricity demand, charging EVs at their maximum rate may potentially increase the grid stress and lead to overall instability. Furthermore, high-rate charging of EVs can accelerate battery degradation, thereby reducing the service lifespan of the fleet. This study aims to reconcile the conflicting incentives of fleet longevity, short-term profitability, and grid stability by simulating a taxi fleet throughout its lifespan in relation to its charging policies and service conditions. We develop an EV fleet simulator to evaluate the battery degradation due to unpredictable charging and ride demand. Consequently, the impact on the power grid through the charging infrastructure is assessed due to these activities. This simulation utilizes publicly accessible real-world travel data from the NYC taxi dataset. We compare a baseline 80-20 fleet charging policy with a reinforcement learning-based policy designed to prolong the fleet's service life and alleviate grid stress. We monitor grid stress, battery degradation, and profitability over five years and find that our learned policy outperforms the baseline. This simulator enables fleet operators to assess the impact of different charging policies on these indicators to make informed decisions in the future.

eess.SY↗

arXiv2025

Energy-Efficient Joint Offloading and Resource Allocation for Deadline-Constrained Tasks in Multi-Access Edge Computing

This paper addresses the deadline-constrained task offloading and resource allocation problem in multi-access edge computing. We aim to determine where each task is offloaded and processed, as well as corresponding communication and computation resource allocations, to maximize the total saved energy for IoT devices, while considering task deadline and system resource constraints. Especially, our system allows each task to be offloaded to one of its accessible access points (APs) and processed on a server that is not co-located with its offloading AP. We formulate this problem as an Integer Nonlinear Programming problem and show it is NP-Hard. To address this problem, we propose a Graph-Matching-based Approximation Algorithm ($\mathtt{GMA}$), the first approximation algorithm of its kind. $\mathtt{GMA}$ leverages linear relaxation, tripartite graph construction, and a Linear Programming rounding technique. We prove that $\mathtt{GMA}$ is a $\frac{1-\alpha}{2+\epsilon}$-approximation algorithm, where $\epsilon$ is a small positive value, and $\alpha$ ($0$$\le$$\alpha$$<$$1$) is a system parameter that ensures the resource allocated to any task by an AP or a server cannot exceed $\alpha$ times its resource capacity. Experiments show that, in practice, $\mathtt{GMA}$'s energy saving achieves $97\%$ of the optimal value on average.

cs.DC↗

arXiv2025

DesCartes Builder: A Tool to Develop Machine-Learning Based Digital Twins

Digital twins (DTs) are increasingly utilized to monitor, manage, and optimize complex systems across various domains, including civil engineering. A core requirement for an effective DT is to act as a fast, accurate, and maintainable surrogate of its physical counterpart, the physical twin (PT). To this end, machine learning (ML) is frequently employed to (i) construct real-time DT prototypes using efficient reduced-order models (ROMs) derived from high-fidelity simulations of the PT's nominal behavior, and (ii) specialize these prototypes into DT instances by leveraging historical sensor data from the target PT. Despite the broad applicability of ML, its use in DT engineering remains largely ad hoc. Indeed, while conventional ML pipelines often train a single model for a specific task, DTs typically require multiple, task- and domain-dependent models. Thus, a more structured approach is required to design DTs. In this paper, we introduce DesCartes Builder, an open-source tool to enable the systematic engineering of ML-based pipelines for real-time DT prototypes and DT instances. The tool leverages an open and flexible visual data flow paradigm to facilitate the specification, composition, and reuse of ML models. It also integrates a library of parameterizable core operations and ML algorithms tailored for DT design. We demonstrate the effectiveness and usability of DesCartes Builder through a civil engineering use case involving the design of a real-time DT prototype to predict the plastic strain of a structure.

cs.SE↗

arXiv2025

Improving Reinforcement Learning Sample-Efficiency using Local Approximation

In this study, we derive Probably Approximately Correct (PAC) bounds on the asymptotic sample-complexity for RL within the infinite-horizon Markov Decision Process (MDP) setting that are sharper than those in existing literature. The premise of our study is twofold: firstly, the further two states are from each other, transition-wise, the less relevant the value of the first state is when learning the $\epsilon$-optimal value of the second; secondly, the amount of 'effort', sample-complexity-wise, expended in learning the $\epsilon$-optimal value of a state is independent of the number of samples required to learn the $\epsilon$-optimal value of a second state that is a sufficient number of transitions away from the first. Inversely, states within each other's vicinity have values that are dependent on each other and will require a similar number of samples to learn. By approximating the original MDP using smaller MDPs constructed using subsets of the original's state-space, we are able to reduce the sample-complexity by a logarithmic factor to $O(SA \log A)$ timesteps, where $S$ and $A$ are the state and action space sizes. We are able to extend these results to an infinite-horizon, model-free setting by constructing a PAC-MDP algorithm with the aforementioned sample-complexity. We conclude with showing how significant the improvement is by comparing our algorithm against prior work in an experimental setting.

cs.LG↗

arXiv2025

Adaptive Multi-prompt Contrastive Network for Few-shot Out-of-distribution Detection

Out-of-distribution (OOD) detection attempts to distinguish outlier samples to prevent models trained on the in-distribution (ID) dataset from producing unavailable outputs. Most OOD detection methods require many IID samples for training, which seriously limits their real-world applications. To this end, we target a challenging setting: few-shot OOD detection, where {Only a few {\em labeled ID} samples are available.} Therefore, few-shot OOD detection is much more challenging than the traditional OOD detection setting. Previous few-shot OOD detection works ignore the distinct diversity between different classes. In this paper, we propose a novel network: Adaptive Multi-prompt Contrastive Network (AMCN), which adapts the ID-OOD separation boundary by learning inter- and intra-class distribution. To compensate for the absence of OOD and scarcity of ID {\em image samples}, we leverage CLIP, connecting text with images, engineering learnable ID and OOD {\em textual prompts}. Specifically, we first generate adaptive prompts (learnable ID prompts, label-fixed OOD prompts and label-adaptive OOD prompts). Then, we generate an adaptive class boundary for each class by introducing a class-wise threshold. Finally, we propose a prompt-guided ID-OOD separation module to control the margin between ID and OOD prompts. Experimental results show that AMCN outperforms other state-of-the-art works.

cs.CV↗

arXiv2025

Energy-Efficient Real-Time Job Mapping and Resource Management in Mobile-Edge Computing

Mobile-edge computing (MEC) has emerged as a promising paradigm for enabling Internet of Things (IoT) devices to handle computation-intensive jobs. Due to the imperfect parallelization of algorithms for job processing on servers and the impact of IoT device mobility on data communication quality in wireless networks, it is crucial to jointly consider server resource allocation and IoT device mobility during job scheduling to fully benefit from MEC, which is often overlooked in existing studies. By jointly considering job scheduling, server resource allocation, and IoT device mobility, we investigate the deadline-constrained job offloading and resource management problem in MEC with both communication and computation contentions, aiming to maximize the total energy saved for IoT devices. For the offline version of the problem, where job information is known in advance, we formulate it as an Integer Linear Programming problem and propose an approximation algorithm, $\mathtt{LHJS}$, with a constant performance guarantee. For the online version, where job information is only known upon release, we propose a heuristic algorithm, $\mathtt{LBS}$, that is invoked whenever a job is released. Finally, we conduct experiments with parameters from real-world applications to evaluate their performance.

cs.DC↗

arXiv2025

Controlled Reach-avoid Set Computation for Discrete-time Polynomial Systems via Convex Optimization

This paper addresses the computation of controlled reach-avoid sets (CRASs) for discrete-time polynomial systems subject to control inputs. A CRAS is a set encompassing initial states from which there exist control inputs driving the system into a target set while avoiding unsafe sets. However, efficiently computing CRASs remains an open problem, especially for discrete-time systems. In this paper, we propose a novel framework for computing CRASs which takes advantage of a probabilistic perspective. This framework transforms the fundamentally nonlinear problem of computing CRASs into a computationally tractable convex optimization problem. By regarding control inputs as disturbances obeying certain probability distributions, a CRAS can be equivalently treated as a 0-reach-avoid set in the probabilistic sense, which consists of initial states from which the probability of eventually entering the target set while remaining within the safe set is greater than zero. Thus, we can employ the convex optimization method of computing 0-reach-avoid sets to estimate CRASs. Furthermore, inspired by the $\epsilon$-greedy strategy widely used in reinforcement learning, we propose an approach that iteratively updates the aforementioned probability distributions imposed on control inputs to compute larger CRASs. We demonstrate the effectiveness of the proposed method on extensive examples.

eess.SY↗

arXiv2025

Convex Computations for Controlled Safety Invariant Sets of Black-box Discrete-time Dynamical Systems

Identifying controlled safety invariant sets (CSISs) is essential for safety-critical systems. This paper addresses the problem of computing CSISs for black-box discrete-time systems, where the dynamics are unknown and only limited simulation data are available. Traditionally, a CSIS requires that for every state in the set, there exists a control input that keeps the system within the set at the next step. However, enforcing such universal invariance, i.e., requiring the set to remain controlled invariant for all states, is often overly restrictive or impractical for black-box systems. To address this, we introduce the notion of a Probably Approximately Correct (PAC) CSIS, in which, with prescribed confidence, there exists a suitable control input to keep the system within the set at the next step for at least a specified fraction of the states. Our approach leverages barrier functions and scenario optimization, yielding a tractable linear programming method for estimating PAC CSISs. Several illustrative examples demonstrate the effectiveness of the proposed framework.

eess.SY↗