Papers › Modulated Diffusion: Accelerating Generative Modeling with Modulated Quantization

Modulated Diffusion: Accelerating Generative Modeling with Modulated Quantization

18 Jun 2025arXiv:2506.22463archive 2025-07-28

Weizhi Gao, Zhichao Hou, Junqi Yin, Feiyi Wang, Linyu Peng, Xiaorui Liu

Diffusion models have emerged as powerful generative models, but their high computation cost in iterative sampling remains a significant bottleneck. In this work, we present an in-depth and insightful study of state-of-the-art acceleration techniques for diffusion models, including caching and quantization, revealing their limitations in computation error and generation quality. To break these limits, this work introduces Modulated Diffusion (MoDiff), an innovative, rigorous, and principled framework that accelerates generative modeling through modulated quantization and error compensation. MoDiff not only inherents the advantages of existing caching and quantization methods but also serves as a general framework to accelerate all diffusion models. The advantages of MoDiff are supported by solid theoretical insight and analysis. In addition, extensive experiments on CIFAR-10 and LSUN demonstrate that MoDiff significant reduces activation quantization from 8 bits to 3 bits without performance degradation in post-training quantization (PTQ). Our code implementation is available at https://github.com/WeizhiGao/MoDiff.

PaperPDFCodeCode Syntology ranMethod as code

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 32 code samples harvested from 1 repository linked to this paper: 16 ran (of which 5 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 6 where Syntology's instrument failed) · 16 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 32 samples from 1 repository, 16 ran (10 with no instrument failure; 6 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.

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 6 of the 16 samples that ran have no recorded date.

WeizhiGao/MoDiff officialmentioned in paperpytorchSyntology: 32 samples, 16 ran (6 instrument-failed) · 16 unverified 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

32 samples harvested: 16 ran (of which 5 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 6 where Syntology's instrument failed) · 16 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
5ran · our draft was wrong
1ran; the contract check could not run
9ran
16unverified

Licence: 32 of the 32 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 WeizhiGao/MoDiff. “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”. For 2 samples that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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.

WeizhiGao/MoDiff official repository 32 samples, 16 ran (6 instrument-failed) · 16 unverified
BasicTransformerBlock in 22 papers qdiff/quant_model.py ran · metamorphic tier: deterministic no licence file found · pointer only · fe821a135405b676 · report fixture: returned BasicTransformerBlock · level recorded 2026-09-22
CrossAttention in 85 papers qdiff/quant_model.py ran · metamorphic tier: deterministic no licence file found · pointer only · edbc6b070aba0b51 · report fixture: returned CrossAttention · level recorded 2026-09-22
CrossQKMatMul qdiff/quant_model.py ran fingerprinted no licence file found · pointer only · 389bb673f605c6d2 · report fixture: signature q:Tensor/3/float, k:Tensor/3/float · returned Tensor · level recorded 2026-09-24
CrossSMVMatMul qdiff/quant_model.py ran no licence file found · pointer only · 9ec250f0fac4b765 · report fixture: signature attn:Tensor/3/float, v:Tensor/3/float · returned Tensor · level recorded 2026-09-24
FeedForward in 198 papers qdiff/quant_model.py ran · metamorphic tier: deterministic no licence file found · pointer only · 3abfca8077fe9d76 · report fixture: returned FeedForward · level recorded 2026-09-22
QKVAttention in 25 papers qdiff/quant_model.py ran no licence file found · pointer only · 941d1452fb38ca5e · report fixture: signature qkv:Tensor/3/float · returned Tensor · level recorded 2026-09-24
QKVAttentionLegacy in 20 papers qdiff/quant_model.py ran · metamorphic tier: deterministic no licence file found · pointer only · 90bafd5717a0d799 · report fixture: returned QKVAttentionLegacy · level recorded 2026-09-22
SMVMatMul qdiff/quant_model.py ran fingerprinted no licence file found · pointer only · a3c010461a312c81 · report fixture: signature weight:Tensor/3/float, v:Tensor/3/float · returned Tensor · level recorded 2026-09-24
UniformAffineQuantizer in 11 papers qdiff/quant_model.py ran · metamorphic tier: deterministic fingerprinted no licence file found · pointer only · d88476b54e8ec555 · report fixture: returned UniformAffineQuantizer · level recorded 2026-09-22
cross_attn_forward qdiff/quant_model.py ran · our draft was wrong no licence file found · pointer only · 86ac950feea0e0fd · report fixture: signature self:callable/0/?, x:Tensor/3/float, context:Tensor/3/float, mask:Tensor/2/bool · returned Tensor · level recorded 2026-09-22
custom_to_np in 6 papers scripts/sample_diffusion_ldm.py ran · our draft was wrong fingerprinted no licence file found · pointer only · 834bbad6b027b869 · report fixture: signature x:Tensor/4/float · returned Tensor · level recorded, date not recorded
custom_to_pil in 9 papers scripts/sample_diffusion_ldm.py ran · our draft was wrong no licence file found · pointer only · 1f425be38a5fdeb5 · report fixture: signature x:Tensor/3/float · returned Image · level recorded, date not recorded
dict2namespace in 31 papers scripts/sample_diffusion_ddim.py ran · our draft was wrong no licence file found · pointer only · 584a23812bb846a3 · report fixture: signature config:dict/0/? · returned Namespace · level recorded, date not recorded
get_beta_schedule in 66 papers scripts/sample_diffusion_ddim.py ran · honoured contract no licence file found · pointer only · ca0cea1ca1eeec89 · 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, date not recorded
logs2pil in 6 papers scripts/sample_diffusion_ldm.py ran · our draft was wrong no licence file found · pointer only · 2eaadefc68b9b67f · report fixture: signature logs:dict/0/?, keys:list/1/str · returned dict · level recorded, date not recorded
torch2hwcuint8 in 6 papers scripts/sample_diffusion_ddim.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted no licence file found · pointer only · c27a10ca9f6b75f3 · report fixture: signature x:Tensor/2/float, clip:bool/0/bool · returned Tensor · level recorded, date not recorded
AttentionBlock in 69 papers qdiff/quant_model.py unverified no licence file found · pointer only · df7a4ce615fa6435 · report
AttnBlock in 22 papers qdiff/quant_model.py unverified no licence file found · pointer only · ae77b73557079f18 · report
BaseQuantBlock qdiff/quant_model.py unverified no licence file found · pointer only · 141229083c7d82ac · report
QKMatMul qdiff/quant_model.py unverified no licence file found · pointer only · cdf30b4cf1e20928 · report
QuantAttentionBlock qdiff/quant_model.py unverified no licence file found · pointer only · 6fec188e5dfb1249 · report
QuantAttnBlock qdiff/quant_model.py unverified no licence file found · pointer only · 3767d1fd2725d986 · report
QuantBasicTransformerBlock qdiff/quant_model.py unverified no licence file found · pointer only · aff831fe69321097 · report
QuantModel qdiff/quant_model.py unverified no licence file found · pointer only · a243ad022c69789a · report
QuantModule qdiff/quant_model.py unverified no licence file found · pointer only · 64b7930a97c62f16 · report
QuantQKMatMul qdiff/quant_model.py unverified no licence file found · pointer only · 8c76a82ae15256d9 · report
QuantResBlock qdiff/quant_model.py unverified no licence file found · pointer only · e5fc108495799e81 · report
QuantResnetBlock qdiff/quant_model.py unverified no licence file found · pointer only · 8bfaf266f4ee379f · report
QuantSMVMatMul qdiff/quant_model.py unverified no licence file found · pointer only · e73fa8043b1b13e1 · report
ResnetBlock in 49 papers qdiff/quant_model.py unverified no licence file found · pointer only · 0c239a04ebd27a00 · report
get_specials qdiff/quant_model.py unverified no licence file found · pointer only · 40f7990566ab7d99 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for Modulated Diffusion (MoDiff) (graph method: Modulated Diffusion (MoDiff)). 47 bricks; individually on a fixture: 26 ran (of which 10 constructed an object rather than computing a result; 17 with no instrument failure: 0 honoured, 2 violated, 15 with no contract checked; 9 where Syntology's instrument failed) · 21 unverified. The assembly has not been run end-to-end.

  1. QuantModel entry · level 0 · unverified · a243ad022c69789a
  2. BaseQuantBlock helper · level 0 · unverified · 141229083c7d82ac
  3. BasicTransformerBlock helper · level 1 · ran · fe821a135405b676
  4. QuantAttnBlock helper · level 0 · unverified · 3767d1fd2725d986
  5. QuantBasicTransformerBlock model · level 0 · unverified · aff831fe69321097
  6. QuantModule model · level 0 · unverified · 64b7930a97c62f16
  7. QuantQKMatMul helper · level 0 · unverified · 8c76a82ae15256d9
  8. QuantSMVMatMul helper · level 0 · unverified · e73fa8043b1b13e1
  9. StraightThrough helper · level 1 · ran · 382e4d6d6d5d10b0
  10. get_specials helper · level 0 · unverified · 40f7990566ab7d99
  11. UniformAffineQuantizer model · level 1 · ran · d88476b54e8ec555
  12. CrossAttention helper · level 1 · ran · edbc6b070aba0b51
  13. FeedForward helper · level 1 · ran · 3abfca8077fe9d76
  14. checkpoint helper · level 0 · unverified · 6fe195fcfd10e259
  15. AttnBlock helper · level 0 · unverified · ae77b73557079f18
  16. cross_attn_forward helper · level 1 · ran · our draft was wrong · 86ac950feea0e0fd
  17. AttentionBlock helper · level 0 · unverified · df7a4ce615fa6435
  18. QKMatMul helper · level 0 · unverified · cdf30b4cf1e20928
  19. QuantAttentionBlock model · level 0 · unverified · 6fec188e5dfb1249
  20. QuantResBlock model · level 0 · unverified · e5fc108495799e81
  21. QuantResnetBlock helper · level 0 · unverified · 8bfaf266f4ee379f
  22. ResBlock helper · level 0 · unverified · 00e230e8cf8490b7
  23. ResnetBlock helper · level 0 · unverified · 0c239a04ebd27a00
  24. SMVMatMul helper · level 1 · ran · a3c010461a312c81
  25. lp_loss helper · level 1 · ran; the contract check could not run · 08c6ea7576220a13
  26. round_ste helper · level 2 · ran · our draft was wrong · 675da8641c602b03
  27. CrossQKMatMul helper · level 1 · ran · 389bb673f605c6d2
  28. CrossSMVMatMul helper · level 1 · ran · 9ec250f0fac4b765
  29. default helper · level 1 · ran · violated contract · 424012cb37b31172
  30. exists helper · level 1 · ran · violated contract · aa5486a3650902d8
  31. max_neg_value helper · level 1 · ran · our draft was wrong · f2fe51a4a5646c6f
  32. GEGLU helper · level 1 · ran · 387673fa5780bf50
  33. CheckpointFunction helper · level 1 · ran · 69360f5b0cc8f069
  34. Normalize helper · level 1 · ran · our draft was wrong · c3a6b977022957cb
  35. QKVAttention helper · level 1 · ran · 941d1452fb38ca5e
  36. QKVAttentionLegacy helper · level 1 · ran · 90bafd5717a0d799
  37. conv_nd helper · level 0 · unverified · fe4eb545bbb728e0
  38. normalization helper · level 1 · ran · our draft was wrong · daafc1c89511b8be
  39. zero_module helper · level 1 · ran · our draft was wrong · 129b804760b3115f
  40. TimestepBlock helper · level 0 · unverified · 7958a01675377ce4
  41. nonlinearity helper · level 1 · ran · our draft was wrong · 3137073275f8c21a
  42. Downsample helper · level 1 · ran · 0feab49c606075bb
  43. Upsample helper · level 1 · ran · 1b4579fe4aec5257
  44. linear helper · level 1 · ran · our draft was wrong · 0e30585f04efffbe
  45. count_flops_attn helper · level 0 · unverified · 6e9c81a619f204c5
  46. GroupNorm32 helper · level 1 · ran · 426c2385fa431d9a
  47. avg_pool_nd helper · level 0 · unverified · ecd0fc28815b65ae

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2506.22463") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Quantization

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Diffusion

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