arXiv subjects

Samuel Mugel

Publications and source records attributed to Samuel Mugel.

At least 19 recordsLinked to original sources

arXiv2026

Where Facts Go Missing: A Layerwise Taxonomy and Per-Layer Attribution of Information Omission in Air-Gapped LLMAgent Pipelines

Air-gapped and on-premises language-model agents can silently omit decision-critical facts at any boundary between source ingestion and final answer generation. We present a nine-layer taxonomy (L0-L8), an instrumented attribution harness, and a conditional omission waterfall that distinguishes deterministic software loss from behavioral non-retrieval. We analyze 75,476 controlled synthetic trials spanning five open-weight model configurations and two inference engines, together with a separate 372-trial real-agent pilot covering FHIR, PubMed, and SEC-EDGAR sources with LangChain and ADK orchestration. The weighted synthetic benchmark yields an omission rate of 0.574 (95% CI: 0.571-0.578); deliberately injected deterministic faults at L0-L3 account for 73.4% of weighted loss under the benchmark allocation. Increasing context length is most strongly associated with omission (odds ratio 7.43, 95% CI: 5.44-10.15). Completed server-profile analyses associate q4 KV cache and scaled RoPE with higher omission. In the real-agent pilot, 57.8% of traces are unsuccessful overall and 50.9% remain unsuccessful after excluding execution errors. These results establish pipeline-level attribution in a controlled stress test, but benchmark allocations, confounded model comparisons, and heuristic behavioral labels do not measure production prevalence or causal architectural effects.

cs.MA↗

arXiv2026

ProvenanceGuard: Source-Aware Factuality Verification for MCP-Based LLM Agents

Tool-using LLM agents increasingly use the Model Context Protocol (MCP) to answer from heterogeneous evidence sources, including search, APIs, databases, clinical records, and formulary tools. Standard factuality metrics usually test whether an answer is supported by pooled evidence, missing a provenance-sensitive failure mode: a claim may be supported somewhere while being attributed to the wrong source. We call this cross-source conflation. We introduce ProvenanceGuard, a source-aware verifier for MCP-grounded answers. It consumes captured MCP traces with stable tool IDs, source IDs, and raw outputs; decomposes answers into atomic claims; routes claims to source-specific evidence; checks support with NLI and a token-alignment proxy; compares stated attribution with the routed source; and returns per-claim verdicts plus an answer-level allow/block decision. Blocked answers can be repaired with retrieval-augmented answer revision and re-verified. We evaluate on 281 medical-domain MCP-agent traces. A 266-trace adjudicated subset yields 2,325 LLM-assisted claim labels split by trace; 361 held-out labels are human-verified. On the 40-trace held-out split, ProvenanceGuard achieves block F1 0.802 and source accuracy 0.858 over 260 source-eligible claims, outperforming source-blind baselines that do not emit claim-to-source IDs. On a harder multi-source benchmark it reaches block F1 0.846, while source-plus-relation accuracy drops to 0.229, showing that exact source ownership remains difficult with semantically close sources. Repair-and-reverify resolves all blocked answers in the full trace set, often via conservative fallback. In 50 controlled clinical conflation probes, ProvenanceGuard detects all injected attribution swaps with no retained wrong attribution. These results show that source attribution is an independent axis for factuality verification in MCP-based agents.

cs.AI↗

arXiv2025

Scaling Laws for Energy Efficiency of Local LLMs

Deploying local large language models and vision-language models on edge devices requires balancing accuracy with constrained computational and energy budgets. Although graphics processors dominate modern artificial-intelligence deployment, most consumer hardware--including laptops, desktops, industrial controllers, and embedded systems--relies on central processing units. Despite this, the computational laws governing central-processing-unit-only inference for local language and vision-language workloads remain largely unexplored. We systematically benchmark large language and vision-language models on two representative central-processing-unit tiers widely used for local inference: a MacBook Pro M2, reflecting mainstream laptop-class deployment, and a Raspberry Pi 5, representing constrained, low-power embedded settings. Using a unified methodology based on continuous sampling of processor and memory usage together with area-under-curve integration, we characterize how computational load scales with input text length for language models and with image resolution for vision-language models. We uncover two empirical scaling laws: (1) computational cost for language-model inference scales approximately linearly with token length; and (2) vision-language models exhibit a preprocessing-driven "resolution knee", where compute remains constant above an internal resolution clamp and decreases sharply below it. Beyond these laws, we show that quantum-inspired compression reduces processor and memory usage by up to 71.9% and energy consumption by up to 62%, while preserving or improving semantic accuracy. These results provide a systematic quantification of multimodal central-processing-unit-only scaling for local language and vision-language workloads, and they identify model compression and input-resolution preprocessing as effective, low-cost levers for sustainable edge inference.

cs.AI↗

arXiv2025

Blockchain Network Analysis using Quantum Inspired Graph Neural Networks & Ensemble Models

In the rapidly evolving domain of financial technology, the detection of illicit transactions within blockchain networks remains a critical challenge, necessitating robust and innovative solutions. This work proposes a novel approach by combining Quantum Inspired Graph Neural Networks (QI-GNN) with flexibility of choice of an Ensemble Model using QBoost or a classic model such as Random Forrest Classifier. This system is tailored specifically for blockchain network analysis in anti-money laundering (AML) efforts. Our methodology to design this system incorporates a novel component, a Canonical Polyadic (CP) decomposition layer within the graph neural network framework, enhancing its capability to process and analyze complex data structures efficiently. Our technical approach has undergone rigorous evaluation against classical machine learning implementations, achieving an F2 score of 74.8% in detecting fraudulent transactions. These results highlight the potential of quantum-inspired techniques, supplemented by the structural advancements of the CP layer, to not only match but potentially exceed traditional methods in complex network analysis for financial security. The findings advocate for a broader adoption and further exploration of quantum-inspired algorithms within the financial sector to effectively combat fraud.

cs.LG↗

arXiv2025

Synthetic Data Generation and Differential Privacy using Tensor Networks' Matrix Product States (MPS)

Synthetic data generation is a key technique in modern artificial intelligence, addressing data scarcity, privacy constraints, and the need for diverse datasets in training robust models. In this work, we propose a method for generating privacy-preserving high-quality synthetic tabular data using Tensor Networks, specifically Matrix Product States (MPS). We benchmark the MPS-based generative model against state-of-the-art models such as CTGAN, VAE, and PrivBayes, focusing on both fidelity and privacy-preserving capabilities. To ensure differential privacy (DP), we integrate noise injection and gradient clipping during training, enabling privacy guarantees via R\'enyi Differential Privacy accounting. Across multiple metrics analyzing data fidelity and downstream machine learning task performance, our results show that MPS outperforms classical models, particularly under strict privacy constraints. This work highlights MPS as a promising tool for privacy-aware synthetic data generation. By combining the expressive power of tensor network representations with formal privacy mechanisms, the proposed approach offers an interpretable and scalable alternative for secure data sharing. Its structured design facilitates integration into sensitive domains where both data quality and confidentiality are critical.

cs.LG↗

arXiv2025

Quantum-Inspired Solver for Simulating Material Deformations

This paper explores the application of tensor networks (TNs) to the simulation of material deformations within the framework of linear elasticity. Material simulations are essential computational tools extensively used in both academic research and industrial applications. TNs, originally developed in quantum mechanics, have recently shown promise in solving partial differential equations (PDEs) due to their potential for exponential speedups over classical algorithms. Our study successfully employs TNs to solve linear elasticity equations with billions of degrees of freedom, achieving exponential reductions in both memory usage and computational time. These results demonstrate the practical viability of TNs as a powerful classical backend for executing quantum-inspired algorithms with significant efficiency gains. This work is based on our research conducted with IKERLAN.

quant-ph↗

arXiv2024

Encoding of Probability Distributions for Quantum Monte Carlo Using Tensor Networks

The application of Tensor Networks (TN) in quantum computing has shown promise, particularly for data loading. However, the assumption that data is readily available often renders the integration of TN techniques into Quantum Monte Carlo (QMC) inefficient, as complete probability distributions would have to be calculated classically. In this paper the tensor-train cross approximation (TT-cross) algorithm is evaluated as a means to address the probability loading problem. We demonstrate the effectiveness of this method on financial distributions, showcasing the TT-cross approach's scalability and accuracy. Our results indicate that the TT-cross method significantly improves circuit depth scalability compared to traditional methods, offering a more efficient pathway for implementing QMC on near-term quantum hardware. The approach also shows high accuracy and scalability in handling high-dimensional financial data, making it a promising solution for quantum finance applications.

quant-ph↗

arXiv2024

CompactifAI: Extreme Compression of Large Language Models using Quantum-Inspired Tensor Networks

Large Language Models (LLMs) such as ChatGPT and LlaMA are advancing rapidly in generative Artificial Intelligence (AI), but their immense size poses significant challenges, such as huge training and inference costs, substantial energy demands, and limitations for on-site deployment. Traditional compression methods such as pruning, distillation, and low-rank approximation focus on reducing the effective number of neurons in the network, while quantization focuses on reducing the numerical precision of individual weights to reduce the model size while keeping the number of neurons fixed. While these compression methods have been relatively successful in practice, there is no compelling reason to believe that truncating the number of neurons is an optimal strategy. In this context, this paper introduces CompactifAI, an innovative LLM compression approach using quantum-inspired Tensor Networks that focuses on the model's correlation space instead, allowing for a more controlled, refined and interpretable model compression. Our method is versatile and can be implemented with - or on top of - other compression techniques. As a benchmark, we demonstrate that a combination of CompactifAI with quantization allows to reduce a 93% the memory size of LlaMA 7B, reducing also 70% the number of parameters, accelerating 50% the training and 25% the inference times of the model, and just with a small accuracy drop of 2% - 3%, going much beyond of what is achievable today by other compression techniques. Our methods also allow to perform a refined layer sensitivity profiling, showing that deeper layers tend to be more suitable for tensor network compression, which is compatible with recent observations on the ineffectiveness of those layers for LLM performance. Our results imply that standard LLMs are, in fact, heavily overparametrized, and do not need to be large at all.

cs.CL↗

arXiv2023

Boosting Defect Detection in Manufacturing using Tensor Convolutional Neural Networks

Defect detection is one of the most important yet challenging tasks in the quality control stage in the manufacturing sector. In this work, we introduce a Tensor Convolutional Neural Network (T-CNN) and examine its performance on a real defect detection application in one of the components of the ultrasonic sensors produced at Robert Bosch's manufacturing plants. Our quantum-inspired T-CNN operates on a reduced model parameter space to substantially improve the training speed and performance of an equivalent CNN model without sacrificing accuracy. More specifically, we demonstrate how T-CNNs are able to reach the same performance as classical CNNs as measured by quality metrics, with up to fifteen times fewer parameters and 4% to 19% faster training times. Our results demonstrate that the T-CNN greatly outperforms the results of traditional human visual inspection, providing value in a current real application in manufacturing.

cs.CV↗

arXiv2023

Application of Tensor Neural Networks to Pricing Bermudan Swaptions

The Cheyette model is a quasi-Gaussian volatility interest rate model widely used to price interest rate derivatives such as European and Bermudan Swaptions for which Monte Carlo simulation has become the industry standard. In low dimensions, these approaches provide accurate and robust prices for European Swaptions but, even in this computationally simple setting, they are known to underestimate the value of Bermudan Swaptions when using the state variables as regressors. This is mainly due to the use of a finite number of predetermined basis functions in the regression. Moreover, in high-dimensional settings, these approaches succumb to the Curse of Dimensionality. To address these issues, Deep-learning techniques have been used to solve the backward Stochastic Differential Equation associated with the value process for European and Bermudan Swaptions; however, these methods are constrained by training time and memory. To overcome these limitations, we propose leveraging Tensor Neural Networks as they can provide significant parameter savings while attaining the same accuracy as classical Dense Neural Networks. In this paper we rigorously benchmark the performance of Tensor Neural Networks and Dense Neural Networks for pricing European and Bermudan Swaptions, and we show that Tensor Neural Networks can be trained faster than Dense Neural Networks and provide more accurate and robust prices than their Dense counterparts.

q-fin.CP↗

arXiv2022

Quantum-Inspired Tensor Neural Networks for Option Pricing

Recent advances in deep learning have enabled us to address the curse of dimensionality (COD) by solving problems in higher dimensions. A subset of such approaches of addressing the COD has led us to solving high-dimensional PDEs. This has resulted in opening doors to solving a variety of real-world problems ranging from mathematical finance to stochastic control for industrial applications. Although feasible, these deep learning methods are still constrained by training time and memory. Tackling these shortcomings, Tensor Neural Networks (TNN) demonstrate that they can provide significant parameter savings while attaining the same accuracy as compared to the classical Dense Neural Network (DNN). In addition, we also show how TNN can be trained faster than DNN for the same accuracy. Besides TNN, we also introduce Tensor Network Initializer (TNN Init), a weight initialization scheme that leads to faster convergence with smaller variance for an equivalent parameter count as compared to a DNN. We benchmark TNN and TNN Init by applying them to solve the parabolic PDE associated with the Heston model, which is widely used in financial pricing theory.

q-fin.PR↗

arXiv2022

Financial Risk Management on a Neutral Atom Quantum Processor

Machine Learning models capable of handling the large datasets collected in the financial world can often become black boxes expensive to run. The quantum computing paradigm suggests new optimization techniques, that combined with classical algorithms, may deliver competitive, faster and more interpretable models. In this work we propose a quantum-enhanced machine learning solution for the prediction of credit rating downgrades, also known as fallen-angels forecasting in the financial risk management field. We implement this solution on a neutral atom Quantum Processing Unit with up to 60 qubits on a real-life dataset. We report competitive performances against the state-of-the-art Random Forest benchmark whilst our model achieves better interpretability and comparable training times. We examine how to improve performance in the near-term validating our ideas with Tensor Networks-based numerical simulations.

quant-ph↗

arXiv2022

Financial Index Tracking via Quantum Computing with Cardinality Constraints

In this work, we demonstrate how to apply non-linear cardinality constraints, important for real-world asset management, to quantum portfolio optimization. This enables us to tackle non-convex portfolio optimization problems using quantum annealing that would otherwise be challenging for classical algorithms. Being able to use cardinality constraints for portfolio optimization opens the doors to new applications for creating innovative portfolios and exchange-traded-funds (ETFs). We apply the methodology to the practical problem of enhanced index tracking and are able to construct smaller portfolios that significantly outperform the risk profile of the target index whilst retaining high degrees of tracking.

quant-ph↗

arXiv2022

Multi-disk clutch optimization using quantum annealing

In this work, we develop a new quantum algorithm to solve a combinatorial problem with significant practical relevance occurring in clutch manufacturing. It is demonstrated how quantum optimization can play a role in real industrial applications in the manufacturing sector. Using the quantum annealer provided by D-Wave Systems, we analyze the performance of the quantum and quantum-classical hybrid solvers and compare them to deterministic- and random-algorithm classical benchmark solvers. The continued evolution of the quantum technology, indicating an expectation for even greater relevance in the future is discussed and the revolutionary potential it could have in the manufacturing sector is highlighted.

quant-ph↗

arXiv2022

Quantum artificial vision for defect detection in manufacturing

In this paper we consider several algorithms for quantum computer vision using Noisy Intermediate-Scale Quantum (NISQ) devices, and benchmark them for a real problem against their classical counterparts. Specifically, we consider two approaches: a quantum Support Vector Machine (QSVM) on a universal gate-based quantum computer, and QBoost on a quantum annealer. The quantum vision systems are benchmarked for an unbalanced dataset of images where the aim is to detect defects in manufactured car pieces. We see that the quantum algorithms outperform their classical counterparts in several ways, with QBoost allowing for larger problems to be analyzed with present-day quantum annealers. Data preprocessing, including dimensionality reduction and contrast enhancement, is also discussed, as well as hyperparameter tuning in QBoost. To the best of our knowledge, this is the first implementation of quantum computer vision systems for a problem of industrial relevance in a manufacturing production line.

quant-ph↗

arXiv2022

Quantum-Inspired Tensor Neural Networks for Partial Differential Equations

Partial Differential Equations (PDEs) are used to model a variety of dynamical systems in science and engineering. Recent advances in deep learning have enabled us to solve them in a higher dimension by addressing the curse of dimensionality in new ways. However, deep learning methods are constrained by training time and memory. To tackle these shortcomings, we implement Tensor Neural Networks (TNN), a quantum-inspired neural network architecture that leverages Tensor Network ideas to improve upon deep learning approaches. We demonstrate that TNN provide significant parameter savings while attaining the same accuracy as compared to the classical Dense Neural Network (DNN). In addition, we also show how TNN can be trained faster than DNN for the same accuracy. We benchmark TNN by applying them to solve parabolic PDEs, specifically the Black-Scholes-Barenblatt equation, widely used in financial pricing theory, empirically showing the advantages of TNN over DNN. Further examples, such as the Hamilton-Jacobi-Bellman equation, are also discussed.

cs.LG↗

arXiv2021

Quantum portfolio value forecasting

We present an algorithm which efficiently estimates the intrinsic long-term value of a portfolio of assets on a quantum computer. The method relies on quantum amplitude estimation to estimate the mean of a novel implementation of the Gordon-Shapiro formula. The choice of loading and readout algorithms makes it possible to price a five-asset portfolio on present day quantum computers, a feat which has not been realised using quantum computing to date. We compare results from two available trapped ion quantum computers. Our results are consistent with classical benchmarks, but result in smaller statistical errors for the same computational cost.

quant-ph↗

arXiv2021

Quantum Portfolio Optimization with Investment Bands and Target Volatility

In this paper we show how to implement in a simple way some complex real-life constraints on the portfolio optimization problem, so that it becomes amenable to quantum optimization algorithms. Specifically, first we explain how to obtain the best investment portfolio with a given target risk. This is important in order to produce portfolios with different risk profiles, as typically offered by financial institutions. Second, we show how to implement individual investment bands, i.e., minimum and maximum possible investments for each asset. This is also important in order to impose diversification and avoid corner solutions. Quite remarkably, we show how to build the constrained cost function as a quadratic binary optimization (QUBO) problem, this being the natural input of quantum annealers. The validity of our implementation is proven by finding the optimal portfolios, using D-Wave Hybrid and its Advantage quantum processor, on portfolios built with all the assets from S&P100 and S&P500. Our results show how practical daily constraints found in quantitative finance can be implemented in a simple way in current NISQ quantum processors, with real data, and under realistic market conditions. In combination with clustering algorithms, our methods would allow to replicate the behaviour of more complex indexes, such as Nasdaq Composite or others, in turn being particularly useful to build and replicate Exchange Traded Funds (ETF).

q-fin.PM↗