Papers › HART: Efficient Visual Generation with Hybrid Autoregressive Transformer

HART: Efficient Visual Generation with Hybrid Autoregressive Transformer

14 Oct 2024arXiv:2410.10812archive 2025-07-28

Haotian Tang, Yecheng Wu, Shang Yang, Enze Xie, Junsong Chen, Junyu Chen, Zhuoyang Zhang, Han Cai, Yao Lu, Song Han

We introduce Hybrid Autoregressive Transformer (HART), an autoregressive (AR) visual generation model capable of directly generating 1024x1024 images, rivaling diffusion models in image generation quality. Existing AR models face limitations due to the poor image reconstruction quality of their discrete tokenizers and the prohibitive training costs associated with generating 1024px images. To address these challenges, we present the hybrid tokenizer, which decomposes the continuous latents from the autoencoder into two components: discrete tokens representing the big picture and continuous tokens representing the residual components that cannot be represented by the discrete tokens. The discrete component is modeled by a scalable-resolution discrete AR model, while the continuous component is learned with a lightweight residual diffusion module with only 37M parameters. Compared with the discrete-only VAR tokenizer, our hybrid approach improves reconstruction FID from 2.11 to 0.30 on MJHQ-30K, leading to a 31% generation FID improvement from 7.85 to 5.38. HART also outperforms state-of-the-art diffusion models in both FID and CLIP score, with 4.5-7.7x higher throughput and 6.9-13.4x lower MACs. Our code is open sourced at https://github.com/mit-han-lab/hart.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 27 code samples harvested from 1 repository linked to this paper: 23 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 6 honoured, 1 violated, 6 with no contract checked; 10 where Syntology's instrument failed) · 4 unverified. Where Syntology's instrument failed: 8 where our contract draft was wrong; 2 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 27 samples from 1 repository, 23 ran (13 with no instrument failure; 10 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 1 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 5 of the 23 samples that ran have no recorded date.

mit-han-lab/hart officialmentioned in papermentioned on GitHubpytorchMITSyntology: 27 samples, 23 ran (10 instrument-failed) · 4 unverified report
FoundationVision/VAR officialmentioned in paperpytorchMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis (arXiv:2412.04431; that paper's own run record: 14 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 1 violated, 3 with no contract checked; 10 where Syntology's instrument failed) · 5 unverified); nothing checks that they implement this paper's method. 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.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

27 samples harvested: 23 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 6 honoured, 1 violated, 6 with no contract checked; 10 where Syntology's instrument failed) · 4 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

6ran · honoured contract
1ran · violated contract
8ran · our draft was wrong
2ran; the contract check could not run
6ran
4unverified

Licence: 4 of the 27 samples are pointer only, meaning Syntology does not serve that copy's text. For 4 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from mit-han-lab/hart. “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. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (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 are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. 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 six 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; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. 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.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

mit-han-lab/hart official repository 27 samples, 23 ran (10 instrument-failed) · 4 unverified
Normalize in 154 papers hart/modules/networks/basic_vae.py ran · our draft was wrong MIT (permissive) · 9fcdaa6e423e8aa7 · report fixture: signature in_channels:int/0/int, num_groups:int/0/int · returned GroupNorm · level recorded 2026-09-22
approx_standard_normal_cdf in 150 papers
census: 4 implementations from 14 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
hart/modules/diffusion/diffusion_utils.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · cfd76fd0d89574a4 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
continuous_gaussian_log_likelihood in 27 papers
census: 2 implementations from 5 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32 · arg 2: rank 2, kind float, dtype float32 · arg 3: rank 2, kind float, dtype float32
hart/modules/diffusion/diffusion_utils.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · ab1c9568b4e13899 · report fixture: signature x:Tensor/2/float, means:Tensor/2/float, log_scales:Tensor/2/float · returned Tensor · level recorded 2026-09-22
drop_path in 325 papers
census: 27 implementations from 61 papers, 5 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 4, kind float, dtype float32
hart/modules/networks/utils.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant MIT (permissive) · 971ae8d2d8313e30 · report fixture: signature x:Tensor/4/float, drop_prob:float/0/float, training:bool/0/bool, scale_by_keep:bool/0/bool · returned Tensor · level recorded 2026-09-22
get_beta_schedule in 66 papers hart/modules/diffusion/gaussian_diffusion.py ran · honoured contract MIT (permissive) here; the stored copy is pointer only · 3e0fa4efc22272d4 · report fixture: signature beta_schedule:str/0/str, beta_start:float/0/float, beta_end:float/0/float, num_diffusion_timesteps:int/0/int · returned ndarray · level recorded 2026-09-22
get_device in 101 papers hart/modules/utils/network.py ran · our draft was wrong MIT (permissive) · 7fe4e244281faf24 · report fixture: signature model:callable/0/? · returned device · level recorded 2026-09-24
get_named_beta_schedule in 119 papers hart/modules/diffusion/gaussian_diffusion.py ran · honoured contract MIT (permissive) · 4f55c34a92359642 · report fixture: signature schedule_name:str/0/str, num_diffusion_timesteps:int/0/int · returned ndarray · level recorded 2026-09-25
get_params_num in 6 papers hart/modules/utils/network.py ran MIT (permissive) · b511bc5bb4607931 · report fixture: signature model:callable/0/?, unit:float/0/float, train_only:bool/0/bool · returned float · level recorded 2026-09-24
get_position_ids_1d hart/modules/networks/basic_hart.py ran MIT (permissive) · 9f89d149e5ae9c22 · report fixture: signature batch_size:int/0/int, L:int/0/int, device:device/0/? · returned Tensor · level recorded 2026-09-24
gumbel_softmax_with_rng in 9 papers hart/modules/networks/utils.py ran MIT (permissive) · e543491ea064c733 · report fixture: signature logits:Tensor/2/float, tau:float/0/float, hard:bool/0/bool, eps:float/0/float, dim:int/0/int, rng:Generator/0/? · returned Tensor · level recorded 2026-09-24
list_mean hart/modules/utils/list.py ran · honoured contract fingerprinted MIT (permissive) · 7daf0d33ee58230c · report fixture: signature x:list/1/Tensor · returned Tensor · level recorded 2026-09-24
list_sum hart/modules/utils/list.py ran · violated contract fingerprinted MIT (permissive) · 09a2b9533add11e1 · report fixture: signature x:list/1/int · returned int · level recorded 2026-09-24
mean_flat in 53 papers
census: 9 implementations from 18 papers, 4 distinct outputs across 3 buckets, one shared input per bucket this sample: bucket 1 of 3, arg 1: rank 3, kind float, dtype float32
hart/modules/diffusion/gaussian_diffusion.py ran · our draft was wrong fingerprinted MIT (permissive) · f6d7c009a8efb8b7 · report fixture: signature tensor:Tensor/3/float · returned Tensor · level recorded, date not recorded
modulate in 137 papers
census: 3 implementations from 12 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32 · arg 2: rank 2, kind float, dtype float32 · arg 3: rank 2, kind float, dtype float32
hart/modules/diffusion/diffloss.py ran · honoured contract fingerprinted MIT (permissive) · 62fcb3912a967a50 · report fixture: signature x:Tensor/2/float, shift:Tensor/2/float, scale:Tensor/2/float · returned Tensor · level recorded, date not recorded
nonlinearity in 140 papers
census: 4 implementations from 11 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
hart/modules/networks/basic_vae.py ran · our draft was wrong fingerprinted MIT (permissive) · 3137073275f8c21a · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
normal_kl in 188 papers
census: 14 implementations from 24 papers, 2 distinct outputs on one shared input
hart/modules/diffusion/diffusion_utils.py ran fingerprinted MIT (permissive) · 5ad3c89a859b3cae · report fixture: signature mean1:Tensor/2/float, logvar1:Tensor/2/float, mean2:Tensor/2/float, logvar2:Tensor/2/float · returned Tensor · level recorded 2026-09-24
rotate_half in 424 papers
census: 24 implementations from 96 papers, 4 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 2, kind float, dtype float32
hart/modules/networks/basic_hart.py ran · our draft was wrong fingerprinted MIT (permissive) · e03d53ba9d4f9ae5 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
sample_with_top_k_top_p_ in 9 papers hart/modules/networks/utils.py ran MIT (permissive) · 80371584448f2f6e · report fixture: signature logits_BlV:Tensor/3/float, top_k:int/0/int, top_p:float/0/float, rng:NoneType/0/?, num_samples:int/0/int · returned Tensor · level recorded 2026-09-24
space_timesteps in 72 papers hart/modules/diffusion/respace.py ran; the contract check could not run (RAISES) · metamorphic tier: deterministic MIT (permissive) here; the stored copy is pointer only · ea9dbc131adf582e · report fixture: signature num_timesteps:int/0/int, section_counts:list/1/int · returned set · level recorded 2026-09-22
torch_randint hart/modules/utils/random.py ran MIT (permissive) · 19f1fd31c903cd91 · report fixture: signature low:int/0/int, high:int/0/int, generator:NoneType/0/? · returned int · level recorded 2026-09-24
torch_random hart/modules/utils/random.py ran · honoured contract MIT (permissive) · ac590aa27c431a22 · report fixture: signature generator:NoneType/0/? · returned float · level recorded 2026-09-24
torch_shuffle hart/modules/utils/random.py ran · our draft was wrong MIT (permissive) · 8829b59194c5fa59 · report fixture: signature src_list:list/1/int, generator:Generator/0/? · returned list · level recorded 2026-09-24
weighted_list_sum hart/modules/utils/list.py ran · honoured contract MIT (permissive) · 48a9a7c6f1af26d2 · report fixture: signature x:list/1/Tensor, weights:list/1/float · returned Tensor · level recorded 2026-09-24
get_position_ids in 6 papers hart/modules/networks/basic_hart.py unverified MIT (permissive) · fda451de9b534788 · report
is_dangerous hart/utils/safety_check.py unverified MIT (permissive) · a2304895c9083cad · report
is_parallel in 15 papers hart/modules/utils/network.py unverified MIT (permissive) · b0d3762fc8297376 · report
make_attn in 22 papers hart/modules/networks/basic_vae.py unverified MIT (permissive) · 793c7e2489dd8dfc · report

Tasks

Image GenerationImage Reconstruction

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Absolute Position EncodingsAdamAttentionBPECLIPDense ConnectionsDiffusionDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

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