Seminars

Tensor networks describe high-dimensional tensors as the contraction of a network (or graph) of low-dimensional tensors. Many interesting tensor can be succinctly represented in this fashion -- from many-body ground states in quantum physics to the...

Understanding deep learning calls for addressing three fundamental questions: expressiveness, optimization and generalization. Expressiveness refers to the ability of compactly sized deep neural networks to represent functions capable of solving...

Chernoff-type bounds study concentration of sums of independent random variables and are extremely useful in various settings. In many settings, the random variables may not be completely independent but only have limited independence. One such...

For any unsatisfiable CNF formula F that is hard to refute in the Resolution proof system, we show that a gadget-composed version of F is hard to refute in any proof system whose lines are computed by efficient communication protocols---or...