Alisa Liu

We introduce the CODAH dataset, an adversarially-constructed evaluation dataset for testing commonsense knowledge using sentence-completion questions that describe situations observed in video. Expand

We introduce a new method that enables definition modeling for multiple senses of the same word from a mixed-sense embedding. Expand

We present a confidence measure that does not require ground truth to estimate separation quality, given a deep model and audio mixture. Expand

Deep generative systems that learn probabilistic models from a corpus of existing music do not explicitly encode knowledge of a musical style, compared to traditional rule-based systems. Expand

We present augmentative generation (Aug-Gen), a method of dataset augmentation for any music generation system trained on a resource-constrained domain. Expand