Papers › The Hessian Penalty: A Weak Prior for Unsupervised Disentanglement

The Hessian Penalty: A Weak Prior for Unsupervised Disentanglement

24 Aug 2020ECCV 2020 8arXiv:2008.10599archive 2025-07-28

William Peebles, John Peebles, Jun-Yan Zhu, Alexei Efros, Antonio Torralba

Existing disentanglement methods for deep generative models rely on hand-picked priors and complex encoder-based architectures. In this paper, we propose the Hessian Penalty, a simple regularization term that encourages the Hessian of a generative model with respect to its input to be diagonal. We introduce a model-agnostic, unbiased stochastic approximation of this term based on Hutchinson's estimator to compute it efficiently during training. Our method can be applied to a wide range of deep generators with just a few lines of code. We show that training with the Hessian Penalty often causes axis-aligned disentanglement to emerge in latent space when applied to ProGAN on several datasets. Additionally, we use our regularization term to identify interpretable directions in BigGAN's latent space in an unsupervised fashion. Finally, we provide empirical evidence that the Hessian Penalty encourages substantial shrinkage when applied to over-parameterized latent spaces.

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="2008.10599")

Code

Syntology Ran 7 of 21 code samples harvested from 1 repository linked to this paper; 14 have no recorded run. Of those that ran: 4 ran · our draft was wrong; 3 ran with no contract checked.

By repository: official repository: 21 samples from 1 repository, 7 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

wpeebles/hessian_penalty officialmentioned in papermentioned on GitHubtfBSD-2-Clause 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

21 samples harvested; 7 ran; 0 honoured the contract we drafted; 14 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

4ran · our draft was wrong
3ran
14unverified

Licence: 0 of the 21 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 wpeebles/hessian_penalty. “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.

D_arch wpeebles/hessian_penalty/biggan_discovery/BigGAN.py official repository ran · our draft was wrong BSD-2-Clause (permissive) · 85eb95cce203dfc7 · report
G_arch wpeebles/hessian_penalty/biggan_discovery/BigGAN.py official repository ran · our draft was wrong BSD-2-Clause (permissive) · e8c308c9283407d7 · report
find_classes wpeebles/hessian_penalty/biggan_discovery/datasets.py official repository ran BSD-2-Clause (permissive) · 52d10a10fb72fb9f · report
gram_schmidt wpeebles/hessian_penalty/biggan_discovery/layers.py official repository ran · our draft was wrong BSD-2-Clause (permissive) · c41836f3e51fa8f1 · report
is_image_file wpeebles/hessian_penalty/biggan_discovery/datasets.py official repository ran BSD-2-Clause (permissive) · 4e6a8faaf8e44abe · report
make_dataset wpeebles/hessian_penalty/biggan_discovery/datasets.py official repository ran BSD-2-Clause (permissive) · 79e61b6269a450e8 · report
rademacher wpeebles/hessian_penalty/hessian_penalty_pytorch.py official repository ran · our draft was wrong BSD-2-Clause (permissive) · 8de600fe2220e241 · report
D_arch wpeebles/hessian_penalty/biggan_discovery/BigGANdeep.py official repository unverified BSD-2-Clause (permissive) · a99a05ac52533be9 · report
G_arch wpeebles/hessian_penalty/biggan_discovery/BigGANdeep.py official repository unverified BSD-2-Clause (permissive) · 44d6a6343f6eb182 · report
fast_gram_schmidt wpeebles/hessian_penalty/biggan_discovery/layers.py official repository unverified BSD-2-Clause (permissive) · 7a6f2360b5090449 · report
get_current_hessian_penalty_loss_weight wpeebles/hessian_penalty/progan_experiments/training/hessian_penalties.py official repository unverified BSD-2-Clause (permissive) · 84a7e309cd2bf8f6 · report
hessian_penalty wpeebles/hessian_penalty/hessian_penalty_np.py official repository unverified BSD-2-Clause (permissive) · 8cd6ed99fbc46c9b · report
hessian_penalty wpeebles/hessian_penalty/hessian_penalty_pytorch.py official repository unverified BSD-2-Clause (permissive) · 3c316ca169bb8f87 · report
loss_dcgan_dis wpeebles/hessian_penalty/biggan_discovery/losses.py official repository unverified BSD-2-Clause (permissive) · 982a26492dc6b7fb · report
loss_dcgan_gen wpeebles/hessian_penalty/biggan_discovery/losses.py official repository unverified BSD-2-Clause (permissive) · 6c34831223553b6f · report
loss_hinge_dis wpeebles/hessian_penalty/biggan_discovery/losses.py official repository unverified BSD-2-Clause (permissive) · 290cd37ab0098d5c · report
multi_layer_second_directional_derivative wpeebles/hessian_penalty/hessian_penalty_pytorch.py official repository unverified BSD-2-Clause (permissive) · c0f4937338732adf · report
multi_layer_second_directional_derivative wpeebles/hessian_penalty/biggan_discovery/hessian_penalty.py official repository unverified BSD-2-Clause (permissive) · eee49efebcd3ff56 · report
proj wpeebles/hessian_penalty/biggan_discovery/layers.py official repository unverified BSD-2-Clause (permissive) · 4015ab1f3ab9b881 · report
rademacher wpeebles/hessian_penalty/biggan_discovery/hessian_penalty.py official repository unverified BSD-2-Clause (permissive) · 58800b4db4b14ffd · report
second_directional_derivative wpeebles/hessian_penalty/biggan_discovery/hessian_penalty.py official repository unverified BSD-2-Clause (permissive) · ea70317070ce2ab5 · report

Tasks

Disentanglement

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

1x1 ConvolutionDense ConnectionsProGANWGAN-GP Loss

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