Search

This chapter is meant to be part of the book "Differential Privacy in Artificial Intelligence: From Theory to Practice" and provides an introduction to Differential Privacy. It starts by illustrating various attempts to protect data privacy, emphasizing where and why they failed, and providing the key desiderata of a robust privacy definition. It then defines the key actors, tasks, and scopes that make up the domain of privacy-preserving data analysis. Following that, it formalizes the definition of Differential Privacy and its inherent properties, including composition, post-processing immunity, and group privacy. The chapter also reviews the basic techniques and mechanisms commonly used to implement Differential Privacy in its pure and approximate forms., Comment: Chapter 1 of book: "Differential Privacy in Artificial Intelligence: From Theory to Practice"

Measurement-based quantum computing offers a promising route towards scalable, universal photonic quantum computation. This approach relies on the deterministic and efficient generation of photonic graph states in which many photons are mutually entangled with various topologies. Recently, deterministic sources of graph states have been demonstrated with quantum emitters in both the optical and microwave domains. In this work, we demonstrate deterministic and reconfigurable graph state generation with optical solid-state integrated quantum emitters. Specifically, we use a single semiconductor quantum dot in a cavity to generate caterpillar graph states, the most general type of graph state that can be produced with a single emitter. By using fast detuned optical pulses, we achieve full control over the spin state, enabling us to vary the entanglement topology at will. We perform quantum state tomography of two successive photons, measuring Bell state fidelities up to 0.80$\pm$0.04 and concurrences up to 0.69$\pm$0.09, while maintaining high photon indistinguishability. This simple optical scheme, compatible with commercially available quantum dot-based single photon sources, brings us a step closer to fault-tolerant quantum computing with spins and photons.

Recent developments in applying machine learning to address Alternating Current Optimal Power Flow (AC OPF) problems have demonstrated significant potential in providing close to optimal solutions for generator dispatch in near real-time. While these learning to optimize methods have demonstrated remarkable performance on steady-state operations, practical applications often demand compliance with dynamic constraints when used for fast-timescale optimization. This paper addresses this gap and develops a real-time stability-constrained OPF model (DynOPF-Net) that simultaneously addresses both optimality and dynamical stability within learning-assisted grid operations. The model is a unique integration of learning to optimize that learns a mapping from load conditions to OPF solutions, capturing the OPF's physical and engineering constraints, with Neural Ordinary Differential Equations, capturing generator dynamics, enabling the inclusion of a subset of stability constraints. Numerical results on the WSCC 9-bus and IEEE 57-bus benchmark systems demonstrate that DynOPF-Net can produce highly accurate AC-OPF solutions while also ensuring system stability, contrasting the unstable results obtained by state-of-the-art LtO methods.

Criminal courts across the United States handle millions of cases every year, and the scheduling of those cases must accommodate a diverse set of constraints, including the preferences and availability of courts, prosecutors, and defense teams. When criminal court schedules are formed, defendants' scheduling preferences often take the least priority, although defendants may face significant consequences (including arrest or detention) for missed court dates. Additionally, studies indicate that defendants' nonappearances impose costs on the courts and other system stakeholders. To address these issues, courts and commentators have begun to recognize that pretrial outcomes for defendants and for the system would be improved with greater attention to court processes, including \emph{court scheduling practices}. There is thus a need for fair criminal court pretrial scheduling systems that account for defendants' preferences and availability, but the collection of such data poses logistical challenges. Furthermore, optimizing schedules fairly across various parties' preferences is a complex optimization problem, even when such data is available. In an effort to construct such a fair scheduling system under data uncertainty, this paper proposes a joint optimization and learning framework that combines machine learning models trained end-to-end with efficient matching algorithms. This framework aims to produce court scheduling schedules that optimize a principled measure of fairness, balancing the availability and preferences of all parties.

The frequency or color of photons is an attractive degree of freedom to encode and distribute the quantum information over long distances. However, the generation of frequency-encoded photonic qubits has so far relied on probabilistic non-linear single-photon sources and inefficient gates. Here, we demonstrate the deterministic generation of photonic qubits hyper-encoded in frequency and polarization based on a semiconductor quantum dot in a cavity. We exploit the double dipole structure of a neutral exciton and demonstrate the generation of any quantum superposition in amplitude and phase, controlled by the polarization of the pump laser pulse. The source generates frequency-polarization single-photon qubits at a rate of 4 MHz corresponding to a generation probability at the first lens of 28 $\pm$ 2%, with a photon number purity > 98%. The photons show an indistinguishability > 91% for each dipole and 88% for a balanced quantum superposition of both. The density matrix of the hyper-encoded photonic state is measured by time-resolved polarization tomography, evidencing a fidelity to the target state of 94 $\pm$ 8% and concurrence of 77 $\pm$ 2%, here limited by frequency overlap in our device. Our approach brings the advantages of quantum dot sources to the field of quantum information processing based on frequency encoding.

Differential equations (DE) constrained optimization plays a critical role in numerous scientific and engineering fields, including energy systems, aerospace engineering, ecology, and finance, where optimal configurations or control strategies must be determined for systems governed by ordinary or stochastic differential equations. Despite its significance, the computational challenges associated with these problems have limited their practical use. To address these limitations, this paper introduces a learning-based approach to DE-constrained optimization that combines techniques from proxy optimization and neural differential equations. The proposed approach uses a dual-network architecture, with one approximating the control strategies, focusing on steady-state constraints, and another solving the associated DEs. This combination enables the approximation of optimal strategies while accounting for dynamic constraints in near real-time. Experiments across problems in energy optimization and finance modeling show that this method provides full compliance with dynamic constraints and it produces results up to 25 times more precise than other methods which do not explicitly model the system's dynamic equations.

The Predict-Then-Optimize framework uses machine learning models to predict unknown parameters of an optimization problem from exogenous features before solving. This setting is common to many real-world decision processes, and recently it has been shown that decision quality can be substantially improved by solving and differentiating the optimization problem within an end-to-end training loop. However, this approach requires significant computational effort in addition to handcrafted, problem-specific rules for backpropagation through the optimization step, challenging its applicability to a broad class of optimization problems. This paper proposes an alternative method, in which optimal solutions are learned directly from the observable features by joint predictive models. The approach is generic, and based on an adaptation of the Learning-to-Optimize paradigm, from which a rich variety of existing techniques can be employed. Experimental evaluations show the ability of several Learning-to-Optimize methods to provide efficient and accurate solutions to an array of challenging Predict-Then-Optimize problems., Comment: arXiv admin note: substantial text overlap with arXiv:2311.13087

Statistical agencies rely on sampling techniques to collect socio-demographic data crucial for policy-making and resource allocation. This paper shows that surveys of important societal relevance introduce sampling errors that unevenly impact group-level estimates, thereby compromising fairness in downstream decisions. To address these issues, this paper introduces an optimization approach modeled on real-world survey design processes, ensuring sampling costs are optimized while maintaining error margins within prescribed tolerances. Additionally, privacy-preserving methods used to determine sampling rates can further impact these fairness issues. The paper explores the impact of differential privacy on the statistics informing the sampling process, revealing a surprising effect: not only the expected negative effect from the addition of noise for differential privacy is negligible, but also this privacy noise can in fact reduce unfairness as it positively biases smaller counts. These findings are validated over an extensive analysis using datasets commonly applied in census statistics.

Speculative decoding has emerged as a widely adopted method to accelerate large language model inference without sacrificing the quality of the model outputs. While this technique has facilitated notable speed improvements by enabling parallel sequence verification, its efficiency remains inherently limited by the reliance on incremental token generation in existing draft models. To overcome this limitation, this paper proposes an adaptation of speculative decoding which uses discrete diffusion models to generate draft sequences. This allows parallelization of both the drafting and verification steps, providing significant speed-ups to the inference process. Our proposed approach, Speculative Diffusion Decoding (SpecDiff), is validated on standard language generation benchmarks and empirically demonstrated to provide a up to 8.7x speed-up over standard generation processes and up to 2.5x speed-up over existing speculative decoding approaches.

Networks are crucial components of many sectors, including telecommunications, healthcare, finance, energy, and transportation.The information carried in such networks often contains sensitive user data, like location data for commuters and packet data for online users. Therefore, when considering data release for networks, one must ensure that data release mechanisms do not leak information about individuals, quantified in a precise mathematical sense. Differential Privacy (DP) is the widely accepted, formal, state-of-the-art technique, which has found use in a variety of real-life settings including the 2020 U.S. Census, Apple users' device data, or Google's location data. Yet, the use of DP comes with new challenges, as the noise added for privacy introduces inaccuracies or biases and further, DP techniques can also distribute these biases disproportionately across different populations, inducing fairness issues. The goal of this paper is to characterize the impact of DP on bias and unfairness in the context of releasing information about networks, taking a departure from previous work which has studied these effects in the context of private population counts release (such as in the U.S. Census). To this end, we consider a network release problem where the network structure is known to all, but the weights on edges must be released privately. We consider the impact of this private release on a simple downstream decision-making task run by a third-party, which is to find the shortest path between any two pairs of nodes and recommend the best route to users. This setting is of highly practical relevance, mirroring scenarios in transportation networks, where preserving privacy while providing accurate routing information is crucial. Our work provides theoretical foundations and empirical evidence into the bias and unfairness arising due to privacy in these networked decision problems., Comment: 32 pages

The goal of future quantum networks is to enable new internet applications that are impossible to achieve using solely classical communication. Up to now, demonstrations of quantum network applications and functionalities on quantum processors have been performed in ad-hoc software that was specific to the experimental setup, programmed to perform one single task (the application experiment) directly into low-level control devices using expertise in experimental physics. Here, we report on the design and implementation of the first architecture capable of executing quantum network applications on quantum processors in platform-independent high-level software. We demonstrate the architecture's capability to execute applications in high-level software, by implementing it as a quantum network operating system -- QNodeOS -- and executing test programs including a delegated computation from a client to a server on two quantum network nodes based on nitrogen-vacancy (NV) centers in diamond. We show how our architecture allows us to maximize the use of quantum network hardware, by multitasking different applications on a quantum network for the first time. Our architecture can be used to execute programs on any quantum processor platform corresponding to our system model, which we illustrate by demonstrating an additional driver for QNodeOS for a trapped-ion quantum network node based on a single $^{40}\text{Ca}^+$ atom. Our architecture lays the groundwork for computer science research in the domain of quantum network programming, and paves the way for the development of software that can bring quantum network technology to society., Comment: 12 pages, 5 figures, supplementary materials (48 pages, 24 figures, 11 tables)

The principle of data minimization aims to reduce the amount of data collected, processed or retained to minimize the potential for misuse, unauthorized access, or data breaches. Rooted in privacy-by-design principles, data minimization has been endorsed by various global data protection regulations. However, its practical implementation remains a challenge due to the lack of a rigorous formulation. This paper addresses this gap and introduces an optimization framework for data minimization based on its legal definitions. It then adapts several optimization algorithms to perform data minimization and conducts a comprehensive evaluation in terms of their compliance with minimization objectives as well as their impact on user privacy. Our analysis underscores the mismatch between the privacy expectations of data minimization and the actual privacy benefits, emphasizing the need for approaches that account for multiple facets of real-world privacy risks.

Low-rank approximation techniques have become the de facto standard for fine-tuning Large Language Models (LLMs) due to their reduced computational and memory requirements. This paper investigates the effectiveness of these methods in capturing the shift of fine-tuning datasets from the initial pre-trained data distribution. Our findings reveal that there are cases in which low-rank fine-tuning falls short in learning such shifts. This, in turn, produces non-negligible side effects, especially when fine-tuning is adopted for toxicity mitigation in pre-trained models, or in scenarios where it is important to provide fair models. Through comprehensive empirical evidence on several models, datasets, and tasks, we show that low-rank fine-tuning inadvertently preserves undesirable biases and toxic behaviors. We also show that this extends to sequential decision-making tasks, emphasizing the need for careful evaluation to promote responsible LLMs development.

Recent work has shown a variety of ways in which machine learning can be used to accelerate the solution of constrained optimization problems. Increasing demand for real-time decision-making capabilities in applications such as artificial intelligence and optimal control has led to a variety of approaches, based on distinct strategies. This work proposes a novel approach to learning optimization, in which the underlying metric space of a proximal operator splitting algorithm is learned so as to maximize its convergence rate. While prior works in optimization theory have derived optimal metrics for limited classes of problems, the results do not extend to many practical problem forms including general Quadratic Programming (QP). This paper shows how differentiable optimization can enable the end-to-end learning of proximal metrics, enhancing the convergence of proximal algorithms for QP problems beyond what is possible based on known theory. Additionally, the results illustrate a strong connection between the learned proximal metrics and active constraints at the optima, leading to an interpretation in which the learning of proximal metrics can be viewed as a form of active set learning.

Learning to Optimize (LtO) is a problem setting in which a machine learning (ML) model is trained to emulate a constrained optimization solver. Learning to produce optimal and feasible solutions subject to complex constraints is a difficult task, but is often made possible by restricting the input space to a limited distribution of related problems. Most LtO methods focus on directly learning solutions to the primal problem, and applying correction schemes or loss function penalties to encourage feasibility. This paper proposes an alternative approach, in which the ML model is trained instead to predict dual solution estimates directly, from which primal estimates are constructed to form dual-feasible solution pairs. This enables an end-to-end training scheme is which the dual objective is maximized as a loss function, and solution estimates iterate toward primal feasibility, emulating a Dual Ascent method. First it is shown that the poor convergence properties of classical Dual Ascent are reflected in poor convergence of the proposed training scheme. Then, by incorporating techniques from practical Augmented Lagrangian methods, we show how the training scheme can be improved to learn highly accurate constrained optimization solvers, for both convex and nonconvex problems.

Many decision processes in artificial intelligence and operations research are modeled by parametric optimization problems whose defining parameters are unknown and must be inferred from observable data. The Predict-Then-Optimize (PtO) paradigm in machine learning aims to maximize downstream decision quality by training the parametric inference model end-to-end with the subsequent constrained optimization. This requires backpropagation through the optimization problem using approximation techniques specific to the problem's form, especially for nondifferentiable linear and mixed-integer programs. This paper extends the PtO methodology to optimization problems with nondifferentiable Ordered Weighted Averaging (OWA) objectives, known for their ability to ensure properties of fairness and robustness in decision models. Through a collection of training techniques and proposed application settings, it shows how optimization of OWA functions can be effectively integrated with parametric prediction for fair and robust optimization under uncertainty.

Learning to Rank (LTR) is one of the most widely used machine learning applications. It is a key component in platforms with profound societal impacts, including job search, healthcare information retrieval, and social media content feeds. Conventional LTR models have been shown to produce biases results, stimulating a discourse on how to address the disparities introduced by ranking systems that solely prioritize user relevance. However, while several models of fair learning to rank have been proposed, they suffer from deficiencies either in accuracy or efficiency, thus limiting their applicability to real-world ranking platforms. This paper shows how efficiently-solvable fair ranking models, based on the optimization of Ordered Weighted Average (OWA) functions, can be integrated into the training loop of an LTR model to achieve favorable balances between fairness, user utility, and runtime efficiency. In particular, this paper is the first to show how to backpropagate through constrained optimizations of OWA objectives, enabling their use in integrated prediction and decision models.

Ensuring privacy-preserving inference on cryptographically secure data is a well-known computational challenge. To alleviate the bottleneck of costly cryptographic computations in non-linear activations, recent methods have suggested linearizing a targeted portion of these activations in neural networks. This technique results in significantly reduced runtimes with often negligible impacts on accuracy. In this paper, we demonstrate that such computational benefits may lead to increased fairness costs. Specifically, we find that reducing the number of ReLU activations disproportionately decreases the accuracy for minority groups compared to majority groups. To explain these observations, we provide a mathematical interpretation under restricted assumptions about the nature of the decision boundary, while also showing the prevalence of this problem across widely used datasets and architectures. Finally, we show how a simple procedure altering the fine-tuning step for linearized models can serve as an effective mitigation strategy., Comment: Extended version of the paper accepted to appear at the Forty-first International Conference on Machine Learning (ICML) 2024

This paper introduces an approach to endow generative diffusion processes the ability to satisfy and certify compliance with constraints and physical principles. The proposed method recast the traditional sampling process of generative diffusion models as a constrained optimization problem, steering the generated data distribution to remain within a specified region to ensure adherence to the given constraints. These capabilities are validated on applications featuring both convex and challenging, non-convex, constraints as well as ordinary differential equations, in domains spanning from synthesizing new materials with precise morphometric properties, generating physics-informed motion, optimizing paths in planning scenarios, and human motion synthesis., Comment: Published at the 38th Conference on Neural Information Processing Systems (NeurIPS 2024)

Quantum emitters such as quantum dots, defects in diamond or in silicon have emerged as efficient single photon sources that are progressively exploited in quantum technologies. In 2019, it was shown that the emitted single photon states often include coherence with the vacuum component. Here we investigate how such photon-number coherence alters quantum interference experiments that are routinely implemented both for characterising or exploiting the generated photons. We show that it strongly modifies intensity correlation measurements in a Hong-Ou-Mandel experiment and leads to errors in indistinguishability estimations. It also results in additional entanglement when performing partial measurements. We illustrate the impact on quantum protocols by evidencing modifications in heralding efficiency and fidelity of two-qubit gates., Comment: 15 pages, 6 figures

The integration of constrained optimization models as components in deep networks has led to promising advances on many specialized learning tasks. A central challenge in this setting is backpropagation through the solution of an optimization problem, which often lacks a closed form. One typical strategy is algorithm unrolling, which relies on automatic differentiation through the entire chain of operations executed by an iterative optimization solver. This paper provides theoretical insights into the backward pass of unrolled optimization, showing that it is asymptotically equivalent to the solution of a linear system by a particular iterative method. Several practical pitfalls of unrolling are demonstrated in light of these insights, and a system called Folded Optimization is proposed to construct more efficient backpropagation rules from unrolled solver implementations. Experiments over various end-to-end optimization and learning tasks demonstrate the advantages of this system both computationally, and in terms of flexibility over various optimization problem forms., Comment: arXiv admin note: text overlap with arXiv:2301.12047

In the machine learning ecosystem, hardware selection is often regarded as a mere utility, overshadowed by the spotlight on algorithms and data. This oversight is particularly problematic in contexts like ML-as-a-service platforms, where users often lack control over the hardware used for model deployment. How does the choice of hardware impact generalization properties? This paper investigates the influence of hardware on the delicate balance between model performance and fairness. We demonstrate that hardware choices can exacerbate existing disparities, attributing these discrepancies to variations in gradient flows and loss surfaces across different demographic groups. Through both theoretical and empirical analysis, the paper not only identifies the underlying factors but also proposes an effective strategy for mitigating hardware-induced performance imbalances.

Many real-world decision processes are modeled by optimization problems whose defining parameters are unknown and must be inferred from observable data. The Predict-Then-Optimize framework uses machine learning models to predict unknown parameters of an optimization problem from features before solving. Recent works show that decision quality can be improved in this setting by solving and differentiating the optimization problem in the training loop, enabling end-to-end training with loss functions defined directly on the resulting decisions. However, this approach can be inefficient and requires handcrafted, problem-specific rules for backpropagation through the optimization step. This paper proposes an alternative method, in which optimal solutions are learned directly from the observable features by predictive models. The approach is generic, and based on an adaptation of the Learning-to-Optimize paradigm, from which a rich variety of existing techniques can be employed. Experimental evaluations show the ability of several Learning-to-Optimize methods to provide efficient, accurate, and flexible solutions to an array of challenging Predict-Then-Optimize problems.

While deep learning gradually penetrates operational planning, its inherent prediction errors may significantly affect electricity prices. This letter examines how prediction errors propagate into electricity prices, revealing notable pricing errors and their spatial disparity in congested power systems. To improve fairness, we propose to embed electricity market-clearing optimization as a deep learning layer. Differentiating through this layer allows for balancing between prediction and pricing errors, as oppose to minimizing prediction errors alone. This layer implicitly optimizes fairness and controls the spatial distribution of price errors across the system. We showcase the price-aware deep learning in the nexus of wind power forecasting and short-term electricity market clearing.

Decision-focused learning (DFL) is an emerging paradigm that integrates machine learning (ML) and constrained optimization to enhance decision quality by training ML models in an end-to-end system. This approach shows significant potential to revolutionize combinatorial decision-making in real-world applications that operate under uncertainty, where estimating unknown parameters within decision models is a major challenge. This paper presents a comprehensive review of DFL, providing an in-depth analysis of both gradient-based and gradient-free techniques used to combine ML and constrained optimization. It evaluates the strengths and limitations of these techniques and includes an extensive empirical evaluation of eleven methods across seven problems. The survey also offers insights into recent advancements and future research directions in DFL. Code and benchmark: https://github.com/PredOpt/predopt-benchmarks, Comment: Experimental Survey and Benchmarking

Quantum computing aims at exploiting quantum phenomena to efficiently perform computations that are unfeasible even for the most powerful classical supercomputers. Among the promising technological approaches, photonic quantum computing offers the advantages of low decoherence, information processing with modest cryogenic requirements, and native integration with classical and quantum networks. To date, quantum computing demonstrations with light have implemented specific tasks with specialized hardware, notably Gaussian Boson Sampling which permitted quantum computational advantage to be reached. Here we report a first user-ready general-purpose quantum computing prototype based on single photons. The device comprises a high-efficiency quantum-dot single-photon source feeding a universal linear optical network on a reconfigurable chip for which hardware errors are compensated by a machine-learned transpilation process. Our full software stack allows remote control of the device to perform computations via logic gates or direct photonic operations. For gate-based computation we benchmark one-, two- and three-qubit gates with state-of-the art fidelities of $99.6\pm0.1 \%$, $93.8\pm0.6 \%$ and $86\pm1.2 \%$ respectively. We also implement a variational quantum eigensolver, which we use to calculate the energy levels of the hydrogen molecule with high accuracy. For photon native computation, we implement a classifier algorithm using a $3$-photon-based quantum neural network and report a first $6$-photon Boson Sampling demonstration on a universal reconfigurable integrated circuit. Finally, we report on a first heralded 3-photon entanglement generation, a key milestone toward measurement-based quantum computing.

In domains with high stakes such as law, recruitment, and healthcare, learning models frequently rely on sensitive user data for inference, necessitating the complete set of features. This not only poses significant privacy risks for individuals but also demands substantial human effort from organizations to verify information accuracy. This paper asks whether it is necessary to use \emph{all} input features for accurate predictions at inference time. The paper demonstrates that, in a personalized setting, individuals may only need to disclose a small subset of their features without compromising decision-making accuracy. The paper also provides an efficient sequential algorithm to determine the appropriate attributes for each individual to provide. Evaluations across various learning tasks show that individuals can potentially report as little as 10\% of their information while maintaining the same accuracy level as a model that employs the full set of user information., Comment: arXiv admin note: substantial text overlap with arXiv:2302.00077

This paper introduces FairDP, a novel mechanism designed to achieve certified fairness with differential privacy (DP). FairDP independently trains models for distinct individual groups, using group-specific clipping terms to assess and bound the disparate impacts of DP. Throughout the training process, the mechanism progressively integrates knowledge from group models to formulate a comprehensive model that balances privacy, utility, and fairness in downstream tasks. Extensive theoretical and empirical analyses validate the efficacy of FairDP and improved trade-offs between model utility, privacy, and fairness compared with existing methods.

The Private Aggregation of Teacher Ensembles (PATE) is a machine learning framework that enables the creation of private models through the combination of multiple "teacher" models and a "student" model. The student model learns to predict an output based on the voting of the teachers, and the resulting model satisfies differential privacy. PATE has been shown to be effective in creating private models in semi-supervised settings or when protecting data labels is a priority. This paper explores whether the use of PATE can result in unfairness, and demonstrates that it can lead to accuracy disparities among groups of individuals. The paper also analyzes the algorithmic and data properties that contribute to these disproportionate impacts, why these aspects are affecting different groups disproportionately, and offers recommendations for mitigating these effects, Comment: This version is a "full version" of the associated IJCAI-23 article. arXiv admin note: substantial text overlap with arXiv:2109.08630

Harnessing the frequency dimension in integrated photonics offers key advantages in terms of scalability, noise resilience, parallelization and compatibility with telecom multiplexing techniques. Integrated ring resonators have been used to generate frequency-entangled states through spontaneous four-wave-mixing. However, state-of-the-art integrated resonators are limited by trade-offs in size, number of frequency modes and spectral separation. We have developed silicon ring resonators with a foot-print below 0.05 mm2 providing more than 70 frequency channels separated by 21 GHz. We exploit the narrow frequency separation to parallelize and independently control 34 single qubit-gates with off-the-shelf electro-optic devices. This allows to fully characterize 17 frequency-bin maximally-entangled qubit pairs by performing quantum state tomography. We demonstrate for the first time a fully connected 5-user quantum network in the frequency domain. These results are a step towards a new generation of quantum circuits implemented with scalable silicon photonics technology, for applications in quantum computing and secure communications.

A number of learning models used in consequential domains, such as to assist in legal, banking, hiring, and healthcare decisions, make use of potentially sensitive users' information to carry out inference. Further, the complete set of features is typically required to perform inference. This not only poses severe privacy risks for the individuals using the learning systems, but also requires companies and organizations massive human efforts to verify the correctness of the released information. This paper asks whether it is necessary to require \emph{all} input features for a model to return accurate predictions at test time and shows that, under a personalized setting, each individual may need to release only a small subset of these features without impacting the final decisions. The paper also provides an efficient sequential algorithm that chooses which attributes should be provided by each individual. Evaluation over several learning tasks shows that individuals may be able to report as little as 10\% of their information to ensure the same level of accuracy of a model that uses the complete users' information.

The remarkable ability of language models (LMs) has also brought challenges at the interface of AI and security. A critical challenge pertains to how much information these models retain and leak about the training data. This is particularly urgent as the typical development of LMs relies on huge, often highly sensitive data, such as emails and chat logs. To contrast this shortcoming, this paper introduces Context-Aware Differentially Private Language Model (CADP-LM) , a privacy-preserving LM framework that relies on two key insights: First, it utilizes the notion of \emph{context} to define and audit the potentially sensitive information. Second, it adopts the notion of Differential Privacy to protect sensitive information and characterize the privacy leakage. A unique characteristic of CADP-LM is its ability to target the protection of sensitive sentences and contexts only, providing a highly accurate private model. Experiments on a variety of datasets and settings demonstrate these strengths of CADP-LM.

Disclosure avoidance (DA) systems are used to safeguard the confidentiality of data while allowing it to be analyzed and disseminated for analytic purposes. These methods, e.g., cell suppression, swapping, and k-anonymity, are commonly applied and may have significant societal and economic implications. However, a formal analysis of their privacy and bias guarantees has been lacking. This paper presents a framework that addresses this gap: it proposes differentially private versions of these mechanisms and derives their privacy bounds. In addition, the paper compares their performance with traditional differential privacy mechanisms in terms of accuracy and fairness on US Census data release and classification tasks. The results show that, contrary to popular beliefs, traditional differential privacy techniques may be superior in terms of accuracy and fairness to differential private counterparts of widely used DA mechanisms.

The integration of constrained optimization models as components in deep networks has led to promising advances on many specialized learning tasks. A central challenge in this setting is backpropagation through the solution of an optimization problem, which typically lacks a closed form. One typical strategy is algorithm unrolling, which relies on automatic differentiation through the operations of an iterative solver. While flexible and general, unrolling can encounter accuracy and efficiency issues in practice. These issues can be avoided by analytical differentiation of the optimization, but current frameworks impose rigid requirements on the optimization problem's form. This paper provides theoretical insights into the backward pass of unrolled optimization, leading to a system for generating efficiently solvable analytical models of backpropagation. Additionally, it proposes a unifying view of unrolling and analytical differentiation through optimization mappings. Experiments over various model-based learning tasks demonstrate the advantages of the approach both computationally and in terms of enhanced expressiveness., Comment: Published in IJCAI

In the framework of optical quantum computing and communications, a major objective consists in building receiving nodes that implement conditional operations on incoming photons, using the interaction with a single stationary qubit. In particular, the quest for scalable nodes motivated the development of cavity-enhanced spin-photon interfaces with solid-state emitters. An important challenge remains, however, to produce a stable, controllable, spin-dependant photon state, in a deterministic way. Here we use a pillar-based high-Q cavity, embedding a singly-charged semiconductor quantum dot, to demonstrate the control of giant polarisation rotations induced by a single electron spin. A complete tomography approach is used to deduce the output polarisation Stokes vector, conditioned by a single spin state. We experimentally demonstrate rotation amplitudes such as $\pm \frac{\pi}{2}$ and $\pi$ in the Poincar\'e sphere, as required for applications based on spin-polarisation mapping and spin-mediated photon-photon gates. In agreement with our modeling, we observe that the environmental noise does not limit the amplitude of the spin-induced rotation, yet slightly degrades the polarisation purity of the output states. We find that the polarisation state of the reflected photons can be manipulated in most of the Poincar\'e sphere, through controlled spin-induced rotations, thanks to moderate cavity birefringence and limited noise. This control allows the operation of spin-photon interfaces in various configurations, including at zero or low magnetic fields, which ensures compatibility with key protocols for photonic cluster state generation.

The remarkable performance of deep learning models and their applications in consequential domains (e.g., facial recognition) introduces important challenges at the intersection of equity and security. Fairness and robustness are two desired notions often required in learning models. Fairness ensures that models do not disproportionately harm (or benefit) some groups over others, while robustness measures the models' resilience against small input perturbations. This paper shows the existence of a dichotomy between fairness and robustness, and analyzes when achieving fairness decreases the model robustness to adversarial samples. The reported analysis sheds light on the factors causing such contrasting behavior, suggesting that distance to the decision boundary across groups as a key explainer for this behavior. Extensive experiments on non-linear models and different architectures validate the theoretical findings in multiple vision domains. Finally, the paper proposes a simple, yet effective, solution to construct models achieving good tradeoffs between fairness and robustness.

Model selection is a strategy aimed at creating accurate and robust models. A key challenge in designing these algorithms is identifying the optimal model for classifying any particular input sample. This paper addresses this challenge and proposes a novel framework for differentiable model selection integrating machine learning and combinatorial optimization. The framework is tailored for ensemble learning, a strategy that combines the outputs of individually pre-trained models, and learns to select appropriate ensemble members for a particular input sample by transforming the ensemble learning task into a differentiable selection program trained end-to-end within the ensemble learning model. Tested on various tasks, the proposed framework demonstrates its versatility and effectiveness, outperforming conventional and advanced consensus rules across a variety of settings and learning tasks., Comment: Full version of the paper appearing in IJCAI-23