As highlighted by recent genome-wide analyses in diverse organisms and cell types,
subcellular targeting of mRNAs has emerged as a major mechanism for cells to establish …

Intracellular targeting of mRNAs has long been recognized as a means to produce proteins
locally, but has only recently emerged as a prevalent mechanism used by a wide variety of …

Scene representation—the process of converting visual sensory data into concise descriptions—is
a requirement for intelligent behavior. Recent work has shown that neural networks …

Membrane‐less organelles, because of their capacity to dynamically, selectively and reversibly
concentrate molecules, are very well adapted for local information processing and rapid …

PatchMatch (PM) is a simple, yet very powerful and successful method for optimizing continuous
labelling problems. The algorithm has two main ingredients: the update of the solution …

We introduce convolutional DRAW, a homogeneous deep generative model achieving state-of-the-art
performance in latent variable image modeling. The algorithm naturally stratifies …

To act and plan in complex environments, we posit that agents should have a mental simulator
of the world with three characteristics: (a) it should build an abstract state representing the …

In this work we introduce a differentiable version of the Compositional Pattern Producing
Network, called the DPPN. Unlike a standard CPPN, the topology of a DPPN is evolved but the …

… not reuse any computations from evaluating p(or|f(o<r,a<r)) for evaluating p(or+1|f(o≤r,a≤r))…
: we allow f to be a recurrent mapping that recursively updates sufficient statistics hr = f(hr−1,…

When agents interact with a complex environment, they must form and maintain beliefs about
the relevant aspects of that environment. We propose a way to efficiently train expressive …