Papers › Very Deep VAEs Generalize Autoregressive Models and Can Outperform Them on Images

Very Deep VAEs Generalize Autoregressive Models and Can Outperform Them on Images

20 Nov 2020ICLR 2021 1arXiv:2011.10650archive 2025-07-28

Rewon Child

We present a hierarchical VAE that, for the first time, generates samples quickly while outperforming the PixelCNN in log-likelihood on all natural image benchmarks. We begin by observing that, in theory, VAEs can actually represent autoregressive models, as well as faster, better models if they exist, when made sufficiently deep. Despite this, autoregressive models have historically outperformed VAEs in log-likelihood. We test if insufficient depth explains why by scaling a VAE to greater stochastic depth than previously explored and evaluating it CIFAR-10, ImageNet, and FFHQ. In comparison to the PixelCNN, these very deep VAEs achieve higher likelihoods, use fewer parameters, generate samples thousands of times faster, and are more easily applied to high-resolution images. Qualitative studies suggest this is because the VAE learns efficient hierarchical visual representations. We release our source code and models at https://github.com/openai/vdvae.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2011.10650")

Code

Syntology Ran 8 of 9 code samples harvested from 4 repositories linked to this paper; 1 has no recorded run. Of those that ran: 2 ran · honoured contract; 3 ran · our draft was wrong; 3 ran · fixture could not drive it.

By repository: official repository: 4 samples from 1 repository, 4 ran; community (archive-listed): 4 samples from 3 repositories, 3 ran; 1 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

openai/vdvae officialmentioned in papermentioned on GitHubpytorch report
ericl122333/latent-vae mentioned on GitHubpytorch report
j-towns/vdvae-jax mentioned on GitHubpytorch report
lupalab/posterior-matching mentioned on GitHubjax report
plai-group/ipa mentioned on GitHubpytorch report
tcl9876/visual-vae mentioned on GitHubjax report
vvvm23/vdvae mentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

9 samples harvested; 8 ran; 2 honoured the contract we drafted; 1 has no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

2ran · honoured contract
3ran · our draft was wrong
3ran · fixture could not drive it
1unverified

Licence: 4 of the 9 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 4 repositories linked to this paper, official or community; each sample names its own and says which. Some samples are identical code Syntology first harvested from another repository; for those, this paper's copy is not located and its licence is not recorded. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

add_vae_arguments openai/vdvae/hps.py official repository ran · our draft was wrong MIT (permissive) · a169b353d342ad7f · report
get_width_settings openai/vdvae/vae.py official repository ran · our draft was wrong fingerprinted MIT (permissive) · 9419abf027f0115d · report
pad_channels openai/vdvae/vae.py official repository ran · fixture could not drive it fingerprinted MIT (permissive) · dd095f22dc8304f0 · report
parse_layer_string openai/vdvae/vae.py official repository ran · our draft was wrong MIT (permissive) · 95d3df9fe1921426 · report
draw_gaussian_diag_samples vvvm23/vdvae/vae.py community (archive-listed) ran · fixture could not drive it no licence file found · pointer only · 31e670f6fff8614d · report
get_evenly_spaced_indices ericl122333/latent-vae/vae/model.py community (archive-listed) ran · honoured contract fingerprinted no licence file found · pointer only · 2ec4359e4ba42b26 · report
get_resolutions ericl122333/latent-vae/vae/model.py community (archive-listed) ran · honoured contract fingerprinted no licence file found · pointer only · 070ddf6b065864d3 · report
reproduce EugenHotaj/pytorch-generative/pytorch_generative/models/vae/vd_vae.py community (archive-listed) unverified MIT (permissive) · 6e32683806d3201d · report
gaussian_analytical_kl identical code first harvested elsewhere ran · fixture could not drive it licence of this copy not recorded · f8752a89ed52495c · report

Tasks

Image Generation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Generation FFHQ 1024 x 1024 Very Deep VAE bits/dimension 2.42 #20 of 20 Archive leaderboard report
Image Generation FFHQ 256 x 256 Very Deep VAE bits/dimension 0.61 #50 of 51 Archive leaderboard report
Image Generation ImageNet 32x32 Very Deep VAE bpd 3.8 #24 of 35 Archive leaderboard report
Image Generation ImageNet 64x64 Very Deep VAE Bits per dim 3.52 #48 of 65 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Hierarchical VAEStochastic Depth

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections