Education
SE(3)-equivariant prediction of molecular wavefunctions and electronic densities
This accuracy makes it possible to derive properties such as energies and forces directly from the wavefunction in an end-to-end manner.
A Spectral Energy Distance for Parallel Speech Synthesis
This spectral energy distance is a proper scoring rule with respect to the distribution over magnitude-spectrograms of the generated waveform audio and offers statistical consistency guarantees.
o - 0.2 0.3 2.8 0.1 1.2 4.3 - 3.4 - 5.2 0.4 0.9 - 0.1 1.6 3.2 0.1 - 0.3 2.1 3.4 2.1 - 0.7 - 0.9 0 2 1 3 4 0 1 2 3 Mean 0.54 2 3 3 1 0 h(x)
While such an update can be accomplished by re-training on the complete data, the process is inefficient and prevents real-time and on-device learning.
Data-Efficient Pipeline for Offline Reinforcement Learning with Limited Data
While this is a common approach in supervised learning, to our knowledge, this has not been discussed in detail in the offline RL setting.