Self-Correcting Non-Chronological Autoregressive Music Generation

We describe a novel approach for generating music using a self-correcting, non-chronological, autoregressive model. We represent music as a sequence of edit events, each of which denotes either the addition or removal of a note—even a note previously generated by the model. During inference, we generate one edit event at a time using direct ancestral sampling. Our method allows the model to fix previous mistakes such as incorrectly sampled notes and prevent the accumulation of errors which autoregressive models are prone to have. Another benefit is a finer, note-bynote control during human and AI collaborative composition. We show through human survey evaluation that our approach generates better results than orderless NADE and Gibbs sampling.