Think of the hidden layers as forming something like a mixed-membership mixture or factor model that’s sprinkled over the units. With this many degrees of freedom, you get lots of different ways to carve up the coefficients without even considering the label-switching problem.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Stan neural network models | 7 | 3705 | May 1, 2018 | |
| Bayesian Neural Networks using R. Neal's code | 21 | 6570 | April 16, 2020 | |
| HMC for Bayesian neural networks? | 7 | 2545 | May 6, 2022 | |
| Multimodality issues in regression model with mixture prior | 4 | 1172 | August 29, 2019 | |
| PolyStan: PolyChord nested sampling and Bayesian evidences for Stan models | 3 | 360 | June 6, 2025 |