As the rate of humanity's data creation increases exponentially with the rise of AI, scientists have been interested in DNA as a way to store digital information. After all, DNA is nature's way of storing data. It encodes genetic information and determines the blueprint of every living thing on earth. And DNA is at least 1,000 times more compact than solid-state hard drives. To demonstrate just how compact, researchers have previously encoded all of Shakespeare's 154 sonnets, 52 pages of Mozart's music, and an episode of the Netflix show "Biohackers" into tiny amounts of DNA.

The National Archives needs help from people with a special set of skills–reading cursive. The archival bureau is seeking volunteer citizen archivists to help them classify and/or transcribe more than 200 years worth of hand-written historical documents. Most of these are from the Revolutionary War-era, known for looped and flowing penmanship. "Reading cursive is a superpower," Suzanne Isaacs, a community manager with the National Archives Catalog told USA Today. "It's not just a matter of whether you learned cursive in school, it's how much you use cursive today."

"Oh my god, this arm is part of me," says Scott Imbrie, who was able to use it to feel objects Two people with paralysis in their hands were able to temporarily regain their sense of touch and feel the shape of objects, thanks to electrical brain stimulation. The approach could one day help people with spinal cord injuries to better carry out everyday activities by controlling a robotic arm that feels like their own. There have been previous efforts to restore touch through brain stimulation, but they were fairly crude.

For years now, brain-computer interfaces (BCI) have incrementally advanced, giving people with spinal injuries or lost limbs the ability to control prosthetics and computer cursors using their signals. But even though the tech has made strides, the replicating subtle, delicate, nuanced sensations of touch has remained just out of reach. Now, however, a team of researchers from the Cortical Bionics Research Group believe they have made a major breakthrough. A pair of patients wearing a BCI was able to control a bionic arm and "feel" tactile edges, shapes, and curvatures along its fingers. The researchers' findings were published today in the journal Science.

In this paper, we develop a family of algorithms for optimizing superposition-structured" or "dirty" statistical estimators for high-dimensional problems involving the minimization of the sum of a smooth loss function with a hybrid regularization. Most of the current approaches are first-order methods, including proximal gradient or Alternating Direction Method of Multipliers (ADMM). We propose a new family of second-order methods where we approximate the loss function using quadratic approximation. The superposition structured regularizer then leads to a subproblem that can be efficiently solved by alternating minimization. We propose a general active subspace selection approach to speed up the solver by utilizing the low-dimensional structure given by the regularizers, and provide convergence guarantees for our algorithm. Empirically, we show that our approach is more than 10 times faster than state-of-the-art first-order approaches for the latent variable graphical model selection problems and multi-task learning problems when there is more than one regularizer. For these problems, our approach appears to be the first algorithm that can extend active subspace ideas to multiple regularizers."

The discovery of neural architectures from simple building blocks is a long-standing goal of Neural Architecture Search (NAS). Hierarchical search spaces are a promising step towards this goal but lack a unifying search space design framework and typically only search over some limited aspect of architectures. In this work, we introduce a unifying search space design framework based on context-free grammars that can naturally and compactly generate expressive hierarchical search spaces that are 100s of orders of magnitude larger than common spaces from the literature. By enhancing and using their properties, we effectively enable search over the complete architecture and can foster regularity. Further, we propose an efficient hierarchical kernel design for a Bayesian Optimization search strategy to efficiently search over such huge spaces.

Matching local visual features is a crucial problem in computer vision and its accuracy greatly depends on the choice of similarity measure. As it is generally very difficult to design by hand a similarity or a kernel perfectly adapted to the data of interest, learning it automatically with as few assumptions as possible is preferable. However, available techniques for kernel learning suffer from several limitations, such as restrictive parametrization or scalability. In this paper, we introduce a simple and flexible family of non-linear kernels which we refer to as Quantized Kernels (QK). QKs are arbitrary kernels in the index space of a data quantizer, i.e., piecewise constant similarities in the original feature space.

The recent availability of large datasets in bio-medicine has inspired the development of representation learning methods for multiple healthcare applications. Despite advances in predictive performance, the clinical utility of such methods is limited when exposed to real-world data. This study develops model diagnostic measures to detect potential pitfalls before deployment without assuming access to external data. Specifically, we focus on modeling realistic data shifts in electrophysiological signals (EEGs) via data transforms and extend the conventional task-based evaluations with analyses of a) the model's latent space and b) predictive uncertainty under these transforms. We conduct experiments on multiple EEG feature encoders and two clinically relevant downstream tasks using publicly available large-scale clinical EEGs.

Recent theoretical work has characterized the dynamics and convergence properties for wide shallow neural networks trained via gradient descent; the asymptotic regime in which the number of parameters tends towards infinity has been dubbed the "mean-field" limit. At initialization, the randomly sampled parameters lead to a deviation from the mean-field limit that is dictated by the classical central limit theorem (CLT). However, the dynamics of training introduces correlations among the parameters raising the question of how the fluctuations evolve during training. Here, we analyze the mean-field dynamics as a Wasserstein gradient flow and prove that the deviations from the mean-field evolution scaled by the width, in the width-asymptotic limit, remain bounded throughout training. This observation has implications for both the approximation rate and the generalization: the upper bound we obtain is controlled by a Monte-Carlo type resampling error, which importantly does not depend on dimension.

Sample efficiency and risk-awareness are central to the development of practical reinforcement learning (RL) for complex decision-making. The former can be addressed by transfer learning, while the latter by optimizing some utility function of the return. However, the problem of transferring skills in a risk-aware manner is not well-understood. In this paper, we address the problem of transferring policies between tasks in a common domain that differ only in their reward functions, in which risk is measured by the variance of reward streams. Our approach begins by extending the idea of generalized policy improvement to maximize entropic utilities, thus extending the dynamic programming's policy improvement operation to sets of policies \emph{and} levels of risk-aversion.