This paper introduces a novel optimization method for differential neural architecture search, based on the theory of prediction with expert advice. Its optimization criterion is well fitted for an architecture-selection, i.e., it minimizes the regret incurred by a sub-optimal selection of operations. Unlike previous search relaxations, that require hard pruning of architectures, our method is designed to dynamically wipe out inferior architectures and enhance superior ones. It achieves an optimal worst-case regret bound and suggests the use of multiple learning-rates, based on the amount of information carried by the backward gradients. Experiments show that our algorithm achieves a strong performance over several image classification datasets. Specifically, it obtains an error rate of 1.6% for CIFAR-10, 23.9% for ImageNet under mobile settings, and achieves state-of-the-art results on three additional datasets.

Yet another indie game is becoming a TV show, as an adaptation of is in the works. This is an action survival RPG that was released in 2025 and won British Game of the Year at the recent BAFTA Game Awards. That's fitting, because is an extremely British game. The Fallout-inspired game is set in the 1960s after a nuclear disaster leads to a quarantine zone being established in the Lake District in northern England. Your character wakes up in a bunker with no recollection of how they got there.

At WIRED Health, immunologist Daniel Davis detailed the ways in which new technologies are enabling a better understanding of the human immune system. The immune system operates at a scale scientists are only just beginning to be able to see. That new view could change how diseases like cancer are tackled. Speaking at WIRED Health on April 16, Daniel Davis, an immunologist at Imperial College London, detailed how researchers are using advanced microscopes to uncover previously invisible dynamics in the human immune system, showing that there are multiple processes happening on a "nanoscale" that was previously out of reach. That new view is already reshaping how immunity is understood.

Combining test statistics from independent trials or experiments is a popular method of meta-analysis. However, there is very limited theoretical understanding of the power of the combined test, especially in high-dimensional models considering composite hypotheses tests. We derive a mathematical framework to study standard meta-analysis testing approaches in the context of the many normal means model, which serves as the platform to investigate more complex models. We introduce a natural and mild restriction on the meta-level combination functions of the local trials. This allows us to mathematically quantify the cost of compressing m trials into real-valued test statistics and combining these. We then derive minimax lower and matching upper bounds for the separation rates of standard combination methods for e.g.

Robotic applications often involve working in environments that are uncertain, dynamic, and partially observable. Recently, diffusion models have been proposed for learning trajectory prediction models trained from expert demonstrations, which can be used for planning in robot tasks. Such models have demonstrated a strong ability to overcome challenges such as multi-modal action distributions, highdimensional output spaces, and training instability. It is crucial to quantify the uncertainty of these dynamics models when using them for planning. In this paper, we quantify the uncertainty of diffusion dynamics models using Conformal Prediction (CP).