Simulating quantum channels is a fundamental primitive in quantum computing, since quantum channels define general (trace-preserving) quantum operations. An arbitrary quantum channel cannot be exactly simulated using a finite-dimensional programmable quantum processor, making it important to develop optimal approximate simulation techniques. In this paper, we study the challenging setting in which the channel to be simulated varies adversarially with time. We propose the use of matrix exponentiated gradient descent (MEGD), an online convex optimization method, and analytically show that it achieves a sublinear regret in time. Through experiments, we validate the main results for time-varying dephasing channels using a programmable generalized teleportation processor.

A large portion of today's world population suffer from vision impairments and wear prescription eyeglasses. However, eyeglasses causes additional bulk and discomfort when used with augmented and virtual reality headsets, thereby negatively impacting the viewer's visual experience. In this work, we remedy the usage of prescription eyeglasses in Virtual Reality (VR) headsets by shifting the optical complexity completely into software and propose a prescription-aware rendering approach for providing sharper and immersive VR imagery. To this end, we develop a differentiable display and visual perception model encapsulating display-specific parameters, color and visual acuity of human visual system and the user-specific refractive errors. Using this differentiable visual perception model, we optimize the rendered imagery in the display using stochastic gradient-descent solvers. This way, we provide prescription glasses-free sharper images for a person with vision impairments. We evaluate our approach on various displays, including desktops and VR headsets, and show significant quality and contrast improvements for users with vision impairments.

In this paper we provide a novel approach to the analysis of kinetic models for label switching, which are used for particle systems that can randomly switch between gradient flows in different energy landscapes. Besides problems in biology and physics, we also demonstrate that stochastic gradient descent, the most popular technique in machine learning, can be understood in this setting, when considering a time-continuous variant. Our analysis is focusing on the case of evolution in a collection of external potentials, for which we provide analytical and numerical results about the evolution as well as the stationary problem.

Modern machine learning requires system designers to specify aspects of the learning pipeline, such as losses, architectures, and optimizers. Meta-learning, or learning-to-learn, instead aims to learn those aspects, and promises to unlock greater capabilities with less manual effort. One particularly ambitious goal of meta-learning is to train general-purpose in-context learning algorithms from scratch, using only black-box models with minimal inductive bias. Such a model takes in training data, and produces test-set predictions across a wide range of problems, without any explicit definition of an inference model, training loss, or optimization algorithm. In this paper we show that Transformers and other blackbox models can be meta-trained to act as general-purpose in-context learners. We characterize transitions between algorithms that generalize, algorithms that memorize, and algorithms that fail to meta-train at all, induced by changes in model size, number of tasks, and meta-optimization. We further show that the capabilities of meta-trained algorithms are bottlenecked by the accessible state size (memory) determining the next prediction, unlike standard models which are thought to be bottlenecked by parameter count. Finally, we propose practical interventions such as biasing the training distribution that improve the meta-training and metageneralization of general-purpose in-context learning algorithms. Meta-learning is the process of automatically discovering new learning algorithms instead of designing them manually (Schmidhuber, 1987). An important quality of human-engineered learning algorithms, such as backpropagation and gradient descent, is their applicability to a wide range of tasks or environments. For learning-to-learn to exceed those capabilities, the meta-learned learning algorithms must be similarily general-purpose. Recently, there has been significant progress toward this goal (Kirsch et al., 2019; Oh et al., 2020). The improved generality of the discovered learning algorithms has been achieved by introducing inductive bias, such as by bottlenecking the architecture or by hiding information, which encourage learning over memorization. Methods include restricting learning rules to use gradients (Metz et al., 2019; Kirsch et al., 2019; Oh et al., 2020), symbolic graphs (Real et al., 2020; Co-Reyes et al., 2021), or parameter sharing (Kirsch & Schmidhuber, 2020; Kirsch et al., 2021). While enabling generalization, these inductive biases come at the cost of increasing the effort to design these systems and potentially restrict the space of discoverable learning algorithms. Instead, we seek to explore general-purpose meta-learning systems with minimal inductive bias.

Moreover, environmental changes due to temperature or mechanical strains can lead to time-varying effects which require adaptive equalization. Adaptive equalizers are indeed commonplace in optical receivers [1, 2], typically implemented via gradient-descent-based least-mean squares filtering [3]. For example, in coherent systems such equalizers can track the inverse Jones matrix of the channel and may also correct for additional distortions such as residual chromatic dispersion [4]. However, the underlying equalizer structure is linear, which limits the type of functionalities that can be expressed and therefore also the performance that can be achieved. To overcome the limitations of linear equalizers, a wide variety of machine learning (ML) algorithms have recently been proposed and verified in hardware (HW). For example, field-programmable gate array (FPGA) implementations of various neural network equalizers were demonstrated for IM/DD links [5], passive optical networks [6], optical interconnects [7], and coherent systems [8].

Many machine learning problems encode their data as a matrix with a possibly very large number of rows and columns. In several applications like neuroscience, image compression or deep reinforcement learning, the principal subspace of such a matrix provides a useful, low-dimensional representation of individual data. Here, we are interested in determining the $d$-dimensional principal subspace of a given matrix from sample entries, i.e. from small random submatrices. Although a number of sample-based methods exist for this problem (e.g. Oja's rule \citep{oja1982simplified}), these assume access to full columns of the matrix or particular matrix structure such as symmetry and cannot be combined as-is with neural networks \citep{baldi1989neural}. In this paper, we derive an algorithm that learns a principal subspace from sample entries, can be applied when the approximate subspace is represented by a neural network, and hence can be scaled to datasets with an effectively infinite number of rows and columns. Our method consists in defining a loss function whose minimizer is the desired principal subspace, and constructing a gradient estimate of this loss whose bias can be controlled. We complement our theoretical analysis with a series of experiments on synthetic matrices, the MNIST dataset \citep{lecun2010mnist} and the reinforcement learning domain PuddleWorld \citep{sutton1995generalization} demonstrating the usefulness of our approach.

The workhorse of machine learning is stochastic gradient descent. To access stochastic gradients, it is common to consider iteratively input/output pairs of a training dataset. Interestingly, it appears that one does not need full supervision to access stochastic gradients, which is the main motivation of this paper. After formalizing the "active labeling" problem, which focuses on active learning with partial supervision, we provide a streaming technique that provably minimizes the ratio of generalization error over the number of samples. We illustrate our technique in depth for robust regression.

Binarized Neural Networks (BNNs) are receiving increasing attention due to their lightweight architecture and ability to run on low-power devices. The state-of-the-art for training classification BNNs restricted to few-shot learning is based on a Mixed Integer Programming (MIP) approach. This paper proposes the BeMi ensemble, a structured architecture of BNNs based on training a single BNN for each possible pair of classes and applying a majority voting scheme to predict the final output. The training of a single BNN discriminating between two classes is achieved by a MIP model that optimizes a lexicographic multi-objective function according to robustness and simplicity principles. This approach results in training networks whose output is not affected by small perturbations on the input and whose number of active weights is as small as possible, while good accuracy is preserved. We computationally validate our model using the MNIST and Fashion-MNIST datasets using up to 40 training images per class. Our structured ensemble outperforms both BNNs trained by stochastic gradient descent and state-of-the-art MIP-based approaches. While the previous approaches achieve an average accuracy of 51.1% on the MNIST dataset, the BeMi ensemble achieves an average accuracy of 61.7% when trained with 10 images per class and 76.4% when trained with 40 images per class.

We study the learning dynamics of self-predictive learning for reinforcement learning, a family of algorithms that learn representations by minimizing the prediction error of their own future latent representations. Despite its recent empirical success, such algorithms have an apparent defect: trivial representations (such as constants) minimize the prediction error, yet it is obviously undesirable to converge to such solutions. Our central insight is that careful designs of the optimization dynamics are critical to learning meaningful representations. We identify that a faster paced optimization of the predictor and semi-gradient updates on the representation, are crucial to preventing the representation collapse. Then in an idealized setup, we show self-predictive learning dynamics carries out spectral decomposition on the state transition matrix, effectively capturing information of the transition dynamics. Building on the theoretical insights, we propose bidirectional self-predictive learning, a novel self-predictive algorithm that learns two representations simultaneously. We examine the robustness of our theoretical insights with a number of small-scale experiments and showcase the promise of the novel representation learning algorithm with large-scale experiments.

Traditional learning-based approaches to student modeling (e.g., predicting grades based on measured activities) generalize poorly to underrepresented/minority student groups due to biases in data availability. In this paper, we propose a Multi-Layer Personalized Federated Learning (MLPFL) methodology which optimizes inference accuracy over different layers of student grouping criteria, such as by course and by demographic subgroups within each course. In our approach, personalized models for individual student subgroups are derived from a global model, which is trained in a distributed fashion via meta-gradient updates that account for subgroup heterogeneity while preserving modeling commonalities that exist across the full dataset. To evaluate our methodology, we consider case studies of two popular downstream student modeling tasks, knowledge tracing and outcome prediction, which leverage multiple modalities of student behavior (e.g., visits to lecture videos and participation on forums) in model training. Experiments on three real-world datasets from online courses demonstrate that our approach obtains substantial improvements over existing student modeling baselines in terms of increasing the average and decreasing the variance of prediction quality across different student subgroups. Visual analysis of the resulting students' knowledge state embeddings confirm that our personalization methodology extracts activity patterns which cluster into different student subgroups, consistent with the performance enhancements we obtain over the baselines.