SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2302.04855 · ICLR · 2023

Trading Information between Latents in Hierarchical Variational Autoencoders

Robert Bamler, Tim Xiao

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 10 functions out of this paper's own repositories and ran 9 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
timxzz/hit canonical 9 of 10
FunctionStatusWhere it lives
ResNet18 Ran timxzz/hit/svhn_classifier.py
code served (permissive licence) · get_code("a7e48b90e47310fe")
beta_pair_to_rgb Ran timxzz/hit/plot_eval.py
code served (permissive licence) · get_code("a18050f6825402cd")
betas_to_rgb Ran timxzz/hit/plot_eval.py
code served (permissive licence) · get_code("6acdec96431c470d")
get_runs_list_from_batch_dir Ran timxzz/hit/utils_eval.py
code served (permissive licence) · get_code("a5fd48f12bc66a01")
get_single_run Ran timxzz/hit/utils_eval.py
code served (permissive licence) · get_code("3eb470d3342c5927")
gumbel_softmax_sample Ran timxzz/hit/utils.py
code served (permissive licence) · get_code("d39a16dc280db7c4")
kl_categorical Ran timxzz/hit/utils.py
code served (permissive licence) · get_code("3042b77f69662558")
mi_bound_given_acc Ran timxzz/hit/plot_eval.py
code served (permissive licence) · get_code("3a062a2eadfae8ba")
sample_gumbel Ran timxzz/hit/utils.py
code served (permissive licence) · get_code("2dde33af7b189247")
load_data Not yet run timxzz/hit/dataloader.py
code served (permissive licence) · get_code("0f7c331072af4dbf")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Variational Autoencoders (VAEs) were originally motivated (Kingma & Welling, 2014) as probabilistic generative models in which one performs approximate Bayesian inference. The proposal of β-VAEs (Higgins et al., 2017) breaks this interpretation and generalizes VAEs to application domains beyond generative modeling (e.g., representation learning, clustering, or lossy data compression) by introducing an objective function that allows practitioners to trade off between the information content ("bit rate") of the latent representation and the distortion of reconstructed data (Alemi et al., 2018). In this paper, we reconsider this rate/distortion trade-off in the context of hierarchical VAEs, i.e., VAEs with more than one layer of latent variables. We identify a general class of inference models for which one can split the rate into contributions from each layer, which can then be tuned independently. We derive theoretical bounds on the performance of downstream tasks as functions of the individual layers' rates and verify our theoretical findings in large-scale experiments. Our results provide guidance for practitioners on which region in rate-space to target for a given application.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2302.04855")
get_code_for_paper("2302.04855")
have("2302.04855")

Connect an agent — have() is free.