Approximating the Predictive Distribution via Adversarially-Trained Hypernetworks

By Christian Henning, Johannes von Oswald, João Sacramento, Simone Carlo Surace, Jean-Pascal Pfister, Benjamin F. Grewe. 2018

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Christian Henning
2. Gyuhak Kim
3. Maria R. Cervera
4. Benjamin Ehret
5. Johannes von Oswald
6. Oscar Chang
7. Neale Ratzlaff
8. Francesco D'Angelo

Citations in the corpus

• Posterior Meta-Replay for Continual Learning (2021, background): “Multiple studies have already investigated the use of weight generators for BNNs [33, 22], ranging from sample-based estimates of the prior-matching term [60] to the use of normalizing flows with shared influence on a set of weights for improved scalability [45].”
• Continual learning with hypernetworks (2019, background): “Such restriction does not apply to hypernetworks, which can in principle model complex multimodal distributions (Louizos & Welling, 2017; Pawlowski et al., 2017; Henning et al., 2018).”
• Principled Weight Initialization for Hypernetworks (2020, background): “…search (Brock et al., 2017; Zhang et al., 2018), Bayesian neural networks (Krueger et al., 2017; Ukai et al., 2018; Pawlowski et al., 2017; Henning et al., 2018; Deutsch et al., 2019), multi-task learning (Pan et al., 2018; Shen et al., 2017; Klocek et al., 2019; Serr ` a et al., 2019;…”
• HyperGAN: A Generative Model for Diverse, Performant Neural Networks (2019, methodology): “(Henning et al., 2018) also use a GAN framework to predict the target weights directly.”