Publications

2025

• Anomaly Detection based on Markov Data: A Statistical Depth Approach
  
  • Clémençon Stéphan
  • Fernández Carlos
  , 2025. The purpose of this article is to extend the notion of statistical depth to the case of sample paths of a Markov chain. Initially introduced to define a center-outward ordering of points in the support of a multivariate distribution, depth functions permit to generalize the notions of quantiles and (signed) ranks for observations in Rd with d > 1, as well as statistical procedures based on such quantities. Here we develop a general theoretical framework for evaluating the depth of a Markov sample path and recovering it statistically from an estimate of its transition probability with (non-) asymptotic guarantees. We also detail some of its applications, focusing particularly on unsupervised anomaly detection. Beyond the theoretical analysis carried out, numerical experiments are displayed, providing empirical evidence of the relevance of the novel concept we introduce here to quantify the degree of abnormality of Markov paths of variable length.
• Adding temporal musical controls on top of pretrained generative models
  
  • Nabi Sarah
  • Demerlé Nils
  • Peeters Geoffroy
  • Bevilacqua Frédéric
  • Esling Philippe
  , 2025. Recent advances in deep generative modeling have enabled high-quality models for musical audio synthesis. However, these approaches remain difficult to control, confined to simple, static attributes and, most importantly, entail retraining a different computationally-heavy architecture for each new control. This is inefficient and impractical as it requires substantial computational resources. In this paper, we propose a novel approach allowing to add time-varying musical controls on top of any pretrained generative models with an exposed latent space (e.g. neural audio codecs), without retraining or finetuning. Our method supports both discrete and continuous attributes by adapting a rectified flow approach with a latent diffusion transformer. We learn an invertible mapping between pretrained latent variables and a new space disentangling explicit control attributes and style variables that capture the remaining factors of variation. This enables both feature extraction from an input, but also editing those features to generate transformed audio samples. Finally, this also introduces the ability to perform synthesis directly from the audio descriptors. We validate our method with 4 datasets going from different musical instruments up to full music recordings, on which we outperform state-of-the-art taskspecific baselines in terms of both generation quality and accuracy of the control by inferring transferred attributes.
• Compositional shield synthesis for safe reinforcement learning in partial observability
  
  • Carr Steven
  • Bakirtzis Georgios
  • Topcu Ufuk
  IEEE Open Journal of Control Systems, 2025, 4, pp.373 - 384. Agents controlled by the output of reinforcement learning (RL) algorithms often transition to unsafe states, particularly in uncertain and partially observable environments. Partially observable Markov decision processes (POMDPs) provide a natural setting for studying such scenarios with limited sensing. Shields filter undesirable actions to ensure safe RL by preserving safety requirements in the agents' policy. However, synthesizing holistic shields is computationally expensive in complex deployment scenarios. We propose the compositional synthesis of shields by modeling safety requirements by parts, thereby improving scalability. In particular, problem formulations in the form of POMDPs using RL algorithms illustrate that an RL agent equipped with the resulting compositional shielding, beyond being safe, converges to higher values of expected reward. By using subproblem formulations, we preserve and improve the ability of shielded agents to require fewer training episodes than unshielded agents, especially in sparse-reward settings. Concretely, we find that compositional shield synthesis allows an RL agent to remain safe in environments two orders of magnitude larger than other state-of-the-art model-based approaches. (10.1109/ojcsys.2025.3611725)
  DOI : 10.1109/ojcsys.2025.3611725
• On the Moreau envelope properties of weakly convex functions
  
  • Renaud Marien
  • Leclaire Arthur
  • Papadakis Nicolas
  , 2025. In this document, we present the main properties satisfied by the Moreau envelope of weakly convex functions. The Moreau envelope has been introduced in convex optimization to regularize convex functionals while preserving their global minimizers. However, the Moreau envelope is also defined for the more general class of weakly convex function and can be a useful tool for optimization in this context. The main properties of the Moreau envelope have been demonstrated for convex functions and are generalized to weakly convex function in various works. This document summarizes the vast literature on the properties of the Moreau envelope and provides the associated proofs.
• Leveraging gem5 and Machine Learning for End-to-End Detection of Cache-based Side-Channel Attack Patterns
  
  • Awais Muhammad
  • Mushtaq Maria
  • Naviner Lirida
  • Bruguier Florent
  • Benoit Pascal
  • Haj-Yahya Jawad
  , 2025. Side-channel attacks (SCAs) exploit physical leakage vectors, including cache states, timing variations, and power consumption in hardware microarchitectures to compromise computational security. This paper introduces an end-to-end, simulation-driven framework that integrates gem5's cycle-accurate architectural models with unsupervised machine learning to automate SCA detection. We simulate Spectre (V1/V2), Prime+Probe, and Flush+Reload attack pattern workloads to generate fine-grained execution traces encompassing BTB access sequences, memory latency distributions, and pipeline stall events. Temporal feature engineering extracts discriminative signatures through branch predictor entropy calculations and miss sequence autocorrelation, while dimensionality reduction via t-SNE optimizes the feature space. These preprocessed traces train an ensemble of Isolation Forest, Variational Autoencoders, and HDBSCAN models to identify attack patterns without predefined templates. Experimental validation demonstrates 0.92 precision and 0.88 recall for cache-based SCAs (Prime+Probe variants), representing a 92% F1-score improvement over SVM baselines. SHAP-based feature attribution reveals BTB-miss run-lengths and memory controller contention queues as critical attack patterns. This pipeline automatically categorize interpretable attack patterns if they are risky for the system, enabling proactive identification of safe and unsafe instructions for secure hardware through gem5's reconfigurable memory hierarchy and cache partitioning mechanisms.
• Overcoming the Technical Hurdles of IoT Adoption: the FITNESS Project Vision and Insights
  
  • Cassiau Nicolas
  • Achir Nadjib
  • Adjih Cédric
  • Andrieux Guillaume
  • Bechkit Walid
  • Ben Hadj Said Siwar
  • Boissier Olivier
  • Bouferroum Aymen
  • Combes Richard
  • Courrèges Fabien
  • Dakdouk Hiba
  • Dhaouadi Amira
  • Diouris Jean-François
  • Elayoubi Salah Eddine
  • Härri Jérôme
  • Kassi Mihia
  • Lagrange Xavier
  • Liu Yandi
  • Loscri Valeria
  • Mannoni Valerian
  • Maudet Sébastien
  • Mitton Nathalie
  • Mokdad Amina
  • Moulay Emmanuel
  • Nadar Ali
  • Nahon Rémi
  • Négrier Romain
  • Nguyen van Tam
  • Pelov Alexander
  • Perrine Clency
  • Pierre Marc
  • Pillement Sébastien
  • Pottier Antony
  • Poulliat Charly
  • Pousset Yannis
  • Rady Mina
  • Sboui Nourhen
  • Sondi Patrick
  • Toutain Laurent
  , 2025. The Internet of Things (IoT) has emerged as a transformative force, enabling seamless connectivity and data exchange between diverse devices and networks. However, the realization of a truly interoperable, secure, and energy-efficient IoT ecosystem remains a significant challenge. In this paper, we present the vision and the first key findings of the FITNESS project, a comprehensive research initiative funded by the French Research Agency (ANR) as part of the France 2030 program. Our work aims to develop solutions that enable the IoT's full potential, emphasizing scalability, interoperability, security, and sustainability, and enabling seamless connectivity and efficient data transmission between diverse IoT devices and networks. We focus in this paper on three critical areas: IoT architecture and interoperability, the place of Artificial Intelligence in IoT, and energy efficiency. Additionally, we identify and analyze key use cases that demonstrate the practical applications of our research, highlighting the importance of real-world implementation to validate and refine our solutions. Through our dedicated research efforts, we have made significant advances across several key areas, while also laying the groundwork for further development in others. These contributions support the emergence of a more secure, efficient, and interoperable IoT ecosystem, and provide a foundation for adoption by stakeholders seeking to harness its transformative potential. (10.5281/zenodo.17119689)
  DOI : 10.5281/zenodo.17119689
• Additivity and chain rules for quantum entropies via multi-index Schatten norms
  
  • Fawzi Omar
  • Kochanowski Jan
  • Rouzé Cambyse
  • van Himbeeck Thomas
  , 2025. The primary entropic measures for quantum states are additive under the tensor product. In the analysis of quantum information processing tasks, the minimum entropy of a set of states, e.g., the minimum output entropy of a channel, often plays a crucial role. A fundamental question in quantum information and cryptography is whether the minimum output entropy remains additive under the tensor product of channels. Here, we establish a general additivity statement for the optimized sandwiched R´enyi entropy of quantum channels. For that, we generalize the results of [Devetak, Junge, King, Ruskai, CMP 2006] to multi-index Schatten norms. As an application, we strengthen the additivity statement of [Van Himbeeck and Brown, 2025] thus allowing the analysis of time-adaptive quantum cryptographic protocols. In addition, we establish chain rules for R´enyi conditional entropies that are similar to the ones used for the generalized entropy accumulation theorem of [Metger, Fawzi, Sutter, Renner, CMP 2024] (10.1007/s00220-026-05567-8)
  DOI : 10.1007/s00220-026-05567-8
• Subgraph Gaussian Embedding Contrast for Self-supervised Graph Representation Learning
  
  • Xie Shifeng
  • Einizade Aref
  • Giraldo Jhony
  , 2026, 16018, pp.430-447. Graph Representation Learning (GRL) is a fundamental task in machine learning, aiming to encode high-dimensional graph-structured data into low-dimensional vectors. Self-Supervised Learning (SSL) methods are widely used in GRL because they can avoid expensive human annotation. In this work, we propose a novel Subgraph Gaussian Embedding Contrast (SubGEC) method. Our approach introduces a subgraph Gaussian embedding module, which adaptively maps subgraphs to a structured Gaussian space, ensuring the preservation of input subgraph characteristics while generating subgraphs with a controlled distribution. We then employ optimal transport distances, more precisely the Wasserstein and Gromov-Wasserstein distances, to effectively measure the similarity between subgraphs, enhancing the robustness of the contrastive learning process. Extensive experiments across multiple benchmarks demonstrate that SubGEC outperforms or presents competitive performance against state-of-the-art approaches. Our findings provide insights into the design of SSL methods for GRL, emphasizing the importance of the distribution of the generated contrastive pairs. (10.1007/978-3-032-06106-5_25)
  DOI : 10.1007/978-3-032-06106-5_25
• T-REX: Table -Refute or Entail eXplainer
  
  • Horstmann Tim Luka
  • Geisenberger Baptiste
  • Alam Mehwish
  , 2025. Verifying textual claims against structured tabular data is a critical yet challenging task in Natural Language Processing with broad real-world impact. While recent advances in Large Language Models (LLMs) have enabled significant progress in table fact-checking, current solutions remain inaccessible to non-experts. We introduce T-REX (Table -Refute or Entail eXplainer), the first live, interactive tool for claim verification over multimodal, multilingual tables using state-of-theart instruction-tuned reasoning LLMs. Designed for accuracy and transparency, T-REX empowers non-experts by providing access to advanced fact-checking technology. The system is openly available online.
• Sports Motion Analysis : From Competition Videos to Data-Driven Interpretations
  
  • Gan Qi
  , 2025. Understanding sports motion is essential for performance analysis and training guidance, and has become a growing research area with the advancement of artificial intelligence (AI) in data analysis. Additionally, due to its analyzable physical mechanisms, sports motion serves as a valuable case study for other motion-related domains, such as pedestrian prediction in autonomous driving, abnormal behavior detection in surveillance, disease diagnosis in medicine, and pose tracking in gaming. However, the black-box nature of modern AI models limits our ability to understand their behavior and decisions. This thesis aims to bridge the gap between sports motion analysis and explainable AI (XAI), leveraging AI's representational power while ensuring interpretability.Two main challenges are addressed: (1) the lack of high-quality sports datasets—despite many online videos, competition footage often suffers from low resolution, motion blur, and noisy backgrounds; and (2) the limited research on interpreting sports motions, with few baseline studies in this area.To tackle these, we focus on long jump, which offers two advantages: the availability of high-quality world-class competition videos and biomechanically well-defined motion sequences suitable for interpretation. To support this work, we built three datasets: (1) Olympic triple jump videos with 2D poses and official distances, (2) World Championship long jump videos with 2D poses and biomechanical features, and (3) long jump videos with 2D poses and jump distances from top level competitions.Our study focuses on two levels of data extraction. First, we estimate athlete poses from video, despite challenges like low frame rates and rare poses. We improve pose accuracy with a post-correction method using 2D sports pose priors, modeled as Neural Distance Fields (NDF) in polar coordinates and trained with a gradient-projection-based augmentation method. Second, we estimate biomechanical features from video using a data-free method, reconstructing athlete trajectories from poses via biomechanical modeling.We explored two paths for interpreting sports motion. The first focuses on biomechanical features: we use a classical interpretable machine learning pipeline by training a quantile random forest regressor and applying SHapley Additive exPlanations (SHAP), Partial Dependence Plots (PDPs), and Individual Conditional Expectation (ICE) plots. This reveals insights aligned with existing literature. However, to capture widespread feature interactions, we propose a new method to estimate not only interaction strengths but also where and how interactions occur.The second path focuses on interpreting pose sequences. We analyze black-box time-series models by generating counterfactual explanations using a sparse autoencoder-based model. Experiments show that this approach yields both faithful and robust explanations, contributing to more interpretable and practical human motion analysis from video data.
• Analytic Rényi Entropy Bounds for Device-Independent Cryptography
  
  • Hahn Thomas
  • Philip Aby
  • Tan Ernest
  • Brown Peter
  , 2025. Device-independent (DI) cryptography represents the highest level of security, enabling cryptographic primitives to be executed safely on uncharacterized devices. Moreover, with successful proof-of-concept demonstrations in randomness expansion, randomness amplification, and quantum key distribution, the field is steadily advancing toward commercial viability. Critical to this continued progression is the development of tighter finite-size security proofs. In this work, we provide a simple method to obtain tighter finite-size security proofs for protocols based on the CHSH game, which is the nonlocality test used in all of the proof-of-concept experiments. We achieve this by analytically solving key-rate optimization problems based on Rényi entropies, providing a simple method to obtain tighter finite-size key rates. (10.48550/arXiv.2507.07365)
  DOI : 10.48550/arXiv.2507.07365
• Execution Platform Contracts
  
  • Bourgeoisat Dorian
  • Kühne Ulrich
  • Brandner Florian
  , 2025. Confidentiality is a crucial security property for many critical applications. As a response to the discovery of numerous micro-architectural side channel attacks such as Spectre, allowing an attacker to extract secret information in pernicious ways, the notion of hardware/software contracts was proposed to formalise the guarantees provided by the hardware to the software. In this paper, we propose to extend this notion to include the guarantees provided by the operating system (OS), so far unspecified in such contracts. We formalize an attacker model adapted to a typical execution model on a shared platform. More precisely, we formalize common thread and memory management policies provided by the OS on top of a hardware model and explore the consequences of potential leaks emerging on such a platform. Our investigation shows that the OS policies play a crucial role in providing security guarantees to code processing sensitive data and thus have to be taken into consideration when writing such code through platform contracts.
• Evict+Spec+Time on RISC-V: Gem5-Based Implementation and Microarchitectural Analysis
  
  • Khan Mahreen
  • Mushtaq Maria
  • Pacalet Renaud
  • Apvrille Ludovic
  , 2025. <div><p>Microarchitectural side-channel attacks are a growing concern and have been widely studied on x86 and ARM architectures, but RISC-V's susceptibility to similar attacks remains understudied. We present the first implementation and evaluation of the Evict+Spec+Time attack on RISC-V, previously demonstrated only on x86 [2]. This advanced variant of Evict+Time integrates three critical phases: eviction, speculation, and timing. First, the attack forcibly evicts target cache lines using RISC-V's cbo.flush instruction via the Zicbom extension [6]. Next, it exploits out-of-order execution to manipulate microarchitectural resources such as the reorder buffer, limiting the processor's ability to mask cache-miss latency. Finally, it infers secret-dependent memory access patterns through precise timing measurements. We validate RISC-V's vulnerability by recovering secret keys from AES T-table implementations. Using the gem5 simulator [4], we provide the first detailed analysis of microarchitectural behavior during the attack, including cache contention, pipeline stalls, and latency variations. These insights establish foundational guidance for developing RISC-V-specific countermeasures against such attacks.</p></div>
• Digital Twin and Digital Thread for System Security and Performance applied to an Electrical Vehicle Charging Use Case
  
  • Heermann Hagen
  • Koch Johannes
  • Grimm Christoph
  • Genius Daniela
  • Apvrille Ludovic
  • Mifdaoui Ahlem
  • Schneider Klaus
  , 2025, pp.1-8. System security requires a solid foundation in both development and operation. During development, performance trade-offs result in security infrastructures that are more or less effective, but usually imperfect. Hence, during operation, runtime monitoring and anomaly detection continuously check for security issues. In this paper, we show how development and operation can be linked. We demonstrate how information and data from development and operation can be aggregated in a digital twin and/or digital thread which is used as the basis for runtime monitoring and anomaly detection. In particular, we address the trade-off between system security and performance in a concrete smart grid system. (10.1109/FDL68117.2025.11165407)
  DOI : 10.1109/FDL68117.2025.11165407
• PESTO: Real‑Time Pitch Estimation with Self‑Supervised Transposition‑Equivariant Objective
  
  • Riou Alain
  • Torres Bernardo
  • Hayes Ben
  • Lattner Stefan
  • Hadjeres Gaëtan
  • Richard Gaël
  • Peeters Geoffroy
  Transactions of the International Society for Music Information Retrieval (TISMIR), Ubiquity Press, 2025, 8 (1), pp.334-352. In this paper, we introduce PESTO, a self-supervised learning approach for single-pitch estimation using a Siamese architecture. Our model processes individual frames of a Variable-Q Transform (VQT) and predicts pitch distributions. The neural network is designed to be equivariant to translations, notably thanks to a Toeplitz fully-connected layer. In addition, we construct pitch-shifted pairs by translating and cropping the VQT frames and train our model with a novel class-based transposition-equivariant objective, eliminating the need for annotated data. Thanks to this architecture and training objective, our model achieves remarkable performance while being very lightweight (130 k parameters). Evaluations on music and speech datasets (MIR-1K, MDB-stem-synth, and PTDB) demonstrate that PESTO not only outperforms self-supervised baselines but also competes with supervised methods, exhibiting superior cross-dataset generalization. Finally, we enhance PESTO's practical utility by developing a streamable VQT implementation using cached convolutions. Combined with our model's low latency (less than 10 ms) and minimal parameter count, this makes PESTO particularly suitable for real-time applications. (10.5334/tismir.251)
  DOI : 10.5334/tismir.251
• A quantitative approach to the GDPR’s anonymisation and “appropriate technical and organisational measures” tests
  
  • Holzenberger Nils
  • Maxwell Winston
  Computer Law & Security Review, Elsevier, 2025, 59, pp.106173-1:106173-13. This article examines two tests from the European General Data Protection Regulation (GDPR): (1) the test for anonymisation (the ''anonymisation test''), and (2) the test for applying ''appropriate technical and organisational measures'' to protect personal data (the ''ATOM test''). Both tests depend on vague legal standards and have given rise to legal disputes and differing interpretations among data protection authorities and courts, including in the context of machine learning. Under the anonymisation test, data are sufficiently anonymised when the risk of identification is ''insignificant'' taking into account ''all means reasonably likely to be used'' by an attacker. Under the ATOM test, measures to protect personal data must be ''appropriate'' with regard to the risks of data loss. Here, we use methods from law and economics to transform these two qualitative tests into quantitative approaches that can be visualized on a graph. For the anonymisation test, we chart different attack efforts and identification probabilities, and propose this as a methodology to help stakeholders discuss what attack efforts are ''reasonably likely'' to be deployed and their likelihood of success. For the ATOM test, we use the Learned Hand formula from law and economics to chart the incremental costs and benefits of privacy protection measures to identify the point where those measures maximize social welfare. The Hand formula permits the negative effects of privacy protection measures, such as the loss of data utility and negative impacts on model fairness, to be taken into account when defining what level of protection is ''appropriate''. We apply our proposed framework to several scenarios, applying the anonymisation test to a Large Language Model, and the ATOM test to a database protected with differential privacy. (10.1016/j.clsr.2025.106173)
  DOI : 10.1016/j.clsr.2025.106173
• Solutions de surveillance avancée des réseaux optiques
  
  • Tomczyk Louis
  , 2025. Cette thèse explore des techniques de traitement du signal appliquées aux réseaux de télécommunications optiques cœur, en se concentrant sur les traitements numériques effectués après la détection. Face à la saturation de la croissance des flux de données mondiaux et aux exigences croissantes en matière de surveillance fine de la couche physique, deux axes de recherche complémentaires sont étudiés. Le premier porte sur la détection et la localisation des pertes de puissance dans les liaisons fibre optique point-à-point. En exploitant la non-commutativité entre la dispersion chromatique et la non-linéarité de Kerr, un estimateur de profil longitudinal de puissance (LPPE) est étudié. Ce travail inclut une analyse théorique de l'interaction entre la dispersion chromatique (CD) et l'auto-modulation de phase (SPM), au cœur des algorithmes LPPE. Le second axe concerne l'égalisation de signaux modulés selon des formats avancés, notamment le modelage probabiliste de constellations (PCS). Inspirée des modèles génératifs, une fonction de coût dérivée des auto encodeurs variationnels (VAE) est intégrée à une unique architecture de filtre adaptatif. Cette approche améliore le suivi de polarisation dans des canaux dynamiques, et permet une correction partielle des erreurs de phase sans recourir à des symboles pilotes.
• When Can Sequence Modelling Approaches Recover the Target Policy In Offline Reinforcement Learning? a Statistical Analysis
  
  • Ghani Abdelghanem
  • Ciblat Philippe
  • Ghogho Mounir
  , 2025. <div><p>We present a theoretical analysis of sample complexity for learning the target policy in offline reinforcement learning (RL) using sequence modeling approaches. Our main theorem establishes bounds on the minimum required number of high-return samples. We identify distinct small-data and largedata regimes, characterized by a critical transition point, and reveal a potential trade-off between context coverage breadth and sampling depth. These findings offer insights into efficient data collection strategies and algorithm design for offline RL.</p></div>
• SHAMaNS: Sound Localization with Hybrid Alpha-Stable Spatial Measure and Neural Steerer
  
  • Di Carlo Diego
  • Fontaine Mathieu
  • Nugraha Aditya Arie
  • Bando Yoshiaki
  • Yoshii Kazuyoshi
  , 2025. This paper describes a sound source localization (SSL) technique that combines an α-stable model for the observed signal with a neural network-based approach for modeling steering vectors. Specifically, a physics-informed neural network, referred to as Neural Steerer, is used to interpolate measured steering vectors (SVs) on a fixed microphone array. This allows for a more robust estimation of the so-called α-stable spatial measure, which represents the most plausible direction of arrival (DOA) of a target signal. As an α-stable model for the non-Gaussian case (α ∈ (0, 2)) theoretically defines a unique spatial measure, we choose to leverage it to account for residual reconstruction error of the Neural Steerer in the downstream tasks. The objective scores indicate that our proposed technique outperforms state-of-the-art methods in the case of multiple sound sources.
• Soft Disentanglement in Frequency Bands for Neural Audio Codecs
  
  • Giniès Benoît
  • Bie Xiaoyu
  • Fercoq Olivier
  • Richard Gaël
  , 2025. In neural-based audio feature extraction, ensuring that representations capture disentangled information is crucial for model interpretability. However, existing disentanglement methods often rely on assumptions that are highly dependent on data characteristics or specific tasks. In this work, we introduce a generalizable approach for learning disentangled features within a neural architecture. Our method applies spectral decomposition to time-domain signals, followed by a multibranch audio codec that operates on the decomposed components. Empirical evaluations demonstrate that our approach achieves better reconstruction and perceptual performance compared to a state-of-the-art baseline while also offering potential advantages for inpainting tasks.
• Age of Information based cache updating with popularity contents: Whittle's index based approach
  
  • Ciblat Philippe
  • Caire Giuseppe
  • Yates Roy D
  , 2025. <div><p>We focus on the scheduling algorithm for updating files from a cloud server to a local server having cache. We consider that only K out of N files can be updated at each timeslot. Each file is time-sensitive and the content relevance is thus measured through the Age of Information. In addition, each file has its own popularity which is time-varying according to a Markovian model. In this paper, we offer two contributions: first, we exhibit Whittle's index for this scheduling problem when the popularity is known and fixed over time. Second, we propose a heuristic based on previous Whittle's index for the timevarying popularity case assuming that only the past popularity is available.</p></div>
• Bayesian Experimental Design with Mutual Information and Learned Errors for Human-Computer Interaction
  
  • Miquel Hugo
  • Gori Julien
  • Rioul Olivier
  , 2025. This work provides a Bayesian framework for handling user errors in interactive systems, with applications in human-computer interaction (HCI) and user modeling. The Bayesian Information Gain (BIG) algorithm [1, 2, 3, 4] is an iterative variant of Bayesian experimental design with mutual information as a cost function, used in HCI. It is a principled approach that maximizes expected information gained from each interaction. More precisely, let Θ be the potential target in the user’s mind with prior distribution p(θ), X be the system feedback, and Y be the corresponding user’s input. In each interaction loop, BIG selects feedback x that maximizes mutual information I(Θ; Y|X= x), assuming a known user model (likelihood) p(y|x,θ), and then updates the posterior distribution p(θ|x,y). This work extends the BIG algorithm to learn from user errors while preserving its mathematical foundations. We incorporate an error rate parameter ϵinto the likelihood function p(y|x,θ,ϵ) and develop an adaptive algorithm that jointly infers both θ and ϵ by updating the posterior p(θ,ϵ|x,y) at each interaction step. We also discuss three simplifying hypotheses for the prior expression p(θ,ϵ) and three user models: (i) zero error; (ii) fixed error rate; (iii) arbitrary random error rate. We prove mathematical continuity between these three models, showing that our adaptive approach naturally extends BIG. We also investigate model mismatch on the overall performance and degradation properties with respect to the standard BIG algorithm. While standard BIG converges quickly with perfect responses, it degrades with even small error rates. The fixed-error model depends critically on correctly estimating the error parameter, while our adaptive model achieves the highest accuracy under varying error conditions, at the expense of additional interactions.
• Causal decompositions of one-dimensional quantum cellular automata
  
  • Vanrietvelde Augustin
  • Mestoudjian Octave
  • Arrighi Pablo
  , 2025. Understanding quantum theory's causal structure stands out as a major matter, since it radically departs from classical notions of causality. We present advances in the research program of causal decompositions, which investigates the existence of an equivalence between the causal and the compositional structures of unitary channels. Our results concern one-dimensional Quantum Cellular Automata (1D QCAs), i.e. unitary channels over a line of N quantum systems (with or without periodic boundary conditions) that feature a causality radius r: a given input cannot causally influence outputs at a distance more than r. We prove that, for N ≥ 4r +1, 1D QCAs all admit causal decompositions: a unitary channel is a 1D QCA if and only if it can be decomposed into a unitary routed circuit of nearest-neighbour interactions, in which its causal structure is compositionally obvious. This provides the first constructive form of 1D QCAs with causality radius one or more, fully elucidating their structure. In addition, we show that this decomposition can be taken to be translation-invariant for the case of translation-invariant QCAs. Our proof of these results makes use of innovative algebraic techniques, leveraging a new framework for capturing partitions into non-factor sub-C* algebras.
• Audio processor parameters: estimating distributions instead of deterministic values
  
  • Peladeau Côme
  • Fourer Dominique
  • Peeters Geoffroy
  , 2025, pp.275-282. Audio effects and sound synthesizers are widely used processors in popular music. Their parameters control the quality of the output sound. Multiple combinations of parameters can lead to the same sound. While recent approaches have been proposed to estimate these parameters given only the output sound, those are deterministic, i.e. they only estimate a single solution among the many possible parameter configurations. In this work, we propose to model the parameters as probability distributions instead of deterministic values. To learn the distributions, we optimize two objectives: (1) we minimize the reconstruction error between the ground truth output sound and the one generated using the estimated parameters, as is it usually done, but also (2) we maximize the parameter diversity, using entropy. We evaluate our approach through two numerical audio experiments to show its effectiveness. These results show how our approach effectively outputs multiple combinations of parameters to match one sound.
• Partitions in quantum theory
  
  • Vanrietvelde Augustin
  • Mestoudjian Octave
  • Arrighi Pablo
  , 2025. Decompositional theories describe the ways in which a global physical system can be split into subsystems, facilitating the study of how different possible partitions of a same system interplay, e.g. in terms of inclusions or signalling. In quantum theory, subsystems are usually framed as sub-C* algebras of the algebra of operators on the global system. However, most decompositional approaches have so far restricted their scope to the case of systems corresponding to factor algebras. We argue that this is a mistake: one should cater for the possibility for non-factor subsystems, arising for instance from symmetry considerations. Building on simple examples, we motivate and present a definition of partitions into an arbitrary number of parts, each of which is a possibly non-factor sub-C* algebra. We discuss its physical interpretation and study its properties, in particular with regards to the structure of algebras' centres. We prove that partitions, defined at the C*-algebraic level, can be represented in terms of a splitting of Hilbert spaces, using the framework of routed quantum circuits. For some partitions, however, such a representation necessarily retains a residual pseudo-nonlocality. We provide an example of this behaviour, given by the partition of a fermionic system into local modes.