Ferromagnetism and superconductivity typically compete with each other since the internal magnetic field generated in a magnet suppresses the formation of spin-singlet Cooper pairs in conventional superconductors. Only a handful of ferromagnetic s...

Initially, it may seem that with the definition of a gradient with respect to the raw threshold, backpropagation and gradient descent could be immediately used to train it. However, just as training weights in a vanilla neural network requires car...

We introduce IndyLSTMs, an independently-recurrent variant of the well-known LSTM recursive neural networks, and compare their accuracy on both academic datasets, and our in-house production datasets.

In this paper, we systematically studied the important but unexplored problem of analyzing differences between widely used network representation learning approaches. We provide a first ever exhaustive evaluation benchmark with 11 datasets and 9 p...

We proposed a method for safe planning under uncertainty and partial observation of teams of heterogeneous robots modelled by MPOMDPs based on barrier functions. We applied our method for safe planning of a team of three robots using high-fidelity...

The projected centre-vortex structure of SU(3) gauge fields is complicated. The qualitative idea of a two-dimensional vortex sheet in four dimensions passing through a three-dimensional visualisation is realised. Perhaps the key new insight is the...

In this paper, we present mdfa, a tool that measures whether a classifier treats differently individuals with similar auditing features but different sensitive attributes. We hope that mdfa’s ability to identify sub-populations with severe violati...