We prove a time hierarchy theorem for the promise-BPTIME. This is considered to be a folklore problem and was thought to follow from the existence of complete problems or through direct diagonalization. We observe that neither argument carries through in some immediate way in the promise version. However, the hierarchy ... more >>>

This work makes two distinct yet related contributions. The first contribution is a new information-theoretic model, the query-with-sketch model, and tools to show lower bounds within it. The second contribution is conceptual, technically builds on the first contribution, and is a barrier in the derandomization of randomized logarithmic space (BPL). ... more >>>

Deep neural networks are the dominant machine learning model. We show that this model is missing a crucial complexity parameter. Today, the standard neural network (NN) model is a circuit whose gates (neurons) are ReLU units. The complexity of a NN is quantified by the depth (number of layers) and ... more >>>