Models released for How do Variational Autoencoders Learn? Insights from Representational Similarity

Bonheme, L. and Grzes, M. (2022) Models released for How do Variational Autoencoders Learn? Insights from Representational Similarity. [Data Collection]

Description

Uncontrolled keywords:	VAE; representational similarity; representation learning; deep learning; variational autoencoders
Subjects:	Q Science > Q Science (General) > Q335 Artificial intelligence Q Science > QA Mathematics (inc Computing science) > QA 76 Software, computer programming, > QA76.87 Neural computers, neural networks
DOI:	10.22024/UniKent/01.01.428
Institutional Unit:	Schools > School of Computing
Former Institutional Unit:	Divisions > Division of Computing, Engineering and Mathematical Sciences > School of Computing
Depositing User:	Lisa Bonheme
Collection period:	From To 1 December 2021 20 May 2022
Last Modified:	03 Jun 2025 08:38
Publication Date:	6 May 2022
URI:	https://data.kent.ac.uk/id/eprint/428
Additional details Additional details Creators: Creators Email ORCID Bonheme, L. lb732@kent.ac.uk https://orcid.org/0000-0002-9166-3971 Grzes, M. m.grzes@kent.ac.uk https://orcid.org/0000-0003-4901-1539 Type of data: Software Collection period: From To 1 December 2021 20 May 2022 Metadata language: English Contact email address: lisa.bonheme@gmail.com Last Modified: 03 Jun 2025 08:38
Actions (login required) View Item

Downloads

Downloads per month over past year

Available Files

Read me

Read me

Read me

README.txt README.txt README.txt Visible to: Anyone Content type: README file Metadata Revision: 2 Mime-Type: text/plain License: GNU All-Permissive License for README files File size: 470B File checksum: MD5 b8c0e33545d85db8e3a8a91cb8a4b3ca

Data

Data

Data

dip_vae_ii.zip dip_vae_ii.zip dip_vae_ii.zip Visible to: Anyone Content type: Data Description: Pre-trained dip-VAE-ii models trained on cars3d, smallnorb, and dsprites with lambda values 1, 2, 5, 10, 20 Metadata Revision: 3 Mime-Type: application/zip License: Creative Commons: Attribution 4.0 (CC-BY) File size: 21GB File checksum: MD5 b50100f7b4dcdc3f1d0c93e6c859f583

beta_vae.zip beta_vae.zip beta_vae.zip Visible to: Anyone Content type: Data Description: The beta VAE models trained on cars3d, smallnorb, and dsprites with beta values 1,2,4,6,8 Metadata Revision: 3 Mime-Type: application/zip License: Creative Commons: Attribution 4.0 (CC-BY) File size: 21GB File checksum: MD5 f4c4274f9d2f61cdc6d8a14c5cb77ebb

beta_tc_vae.zip beta_tc_vae.zip beta_tc_vae.zip Visible to: Anyone Content type: Data Description: Pre-trained beta TC-VAE models trained on cars3d, smallnorb, and dsprites with beta values 1, 2, 4, 6, 8 Metadata Revision: 3 Mime-Type: application/zip License: Creative Commons: Attribution 4.0 (CC-BY) File size: 21GB File checksum: MD5 816f69c200c7f08b0e9306b39f8b4999

annealed_vae.zip annealed_vae.zip annealed_vae.zip Visible to: Anyone Content type: Data Description: Pre-trained Annealed-VAE models trained on cars3d, smallnorb, and dsprites with c_max values 5, 10, 20, 50, 75 Metadata Revision: 2 Mime-Type: application/zip License: Creative Commons: Attribution 4.0 (CC-BY) File size: 20GB File checksum: MD5 0a2c71e2197f7307eb83429fc4d781ac