The Appendix is structured as follows: We provide a proof of conditional guarantees in EENNs for (hard) PoE in Appendix A . We conduct an ablation study for our P A model in Appendix B.2 . We report results of NLP experiments in Appendix B.4 . We discuss anytime regression and deep ensembles in Appendix B.6 . We propose a technique for controlling the violations of conditional monotonicity in P A in Appendix B.8 .

Synthetic face recognition (SFR) aims to generate synthetic face datasets that mimic the distribution of real face data, which allows for training face recognition models in a privacy-preserving manner.

The authors contributed equally to this work 37th Conference on Neural Information Processing Systems (NeurIPS 2023). Developing upper bounds on the generalisation gap, i.e., generalisation bounds has been a longstanding topic in statistical learning.

Unsupervised skill discovery carries the promise that an intelligent agent can learn reusable skills through autonomous, reward-free environment interaction.

Hardware-specific optimizations in machine learning (ML) frameworks can cause numerical deviations of inference results. Quite surprisingly, despite using a fixed trained model and fixed input data, inference results are not consistent across platforms, and sometimes not even deterministic on the same platform. We study the causes of these numerical deviations for convolutional neural networks (CNN) on realistic end-to-end inference pipelines and in isolated experiments. Results from 75 distinct platforms suggest that the main causes of deviations on CPUs are differences in SIMD use, and the selection of convolution algorithms at runtime on GPUs. We link the causes and propagation effects to properties of the ML model and evaluate potential mitigations. We make our research code publicly available.

In this paper, we contribute to both theoretical analyses.