Papers › Controlling Text-to-Image Diffusion by Orthogonal Finetuning

Controlling Text-to-Image Diffusion by Orthogonal Finetuning

12 Jun 2023NeurIPS 2023 11arXiv:2306.07280archive 2025-07-28

Zeju Qiu, Weiyang Liu, Haiwen Feng, Yuxuan Xue, Yao Feng, Zhen Liu, Dan Zhang, Adrian Weller, Bernhard Schölkopf

Large text-to-image diffusion models have impressive capabilities in generating photorealistic images from text prompts. How to effectively guide or control these powerful models to perform different downstream tasks becomes an important open problem. To tackle this challenge, we introduce a principled finetuning method -- Orthogonal Finetuning (OFT), for adapting text-to-image diffusion models to downstream tasks. Unlike existing methods, OFT can provably preserve hyperspherical energy which characterizes the pairwise neuron relationship on the unit hypersphere. We find that this property is crucial for preserving the semantic generation ability of text-to-image diffusion models. To improve finetuning stability, we further propose Constrained Orthogonal Finetuning (COFT) which imposes an additional radius constraint to the hypersphere. Specifically, we consider two important finetuning text-to-image tasks: subject-driven generation where the goal is to generate subject-specific images given a few images of a subject and a text prompt, and controllable generation where the goal is to enable the model to take in additional control signals. We empirically show that our OFT framework outperforms existing methods in generation quality and convergence speed.

PaperPDFConference PDFCodeCode 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 22 code samples harvested from 2 repositories linked to this paper: 12 ran (of which 1 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 5 where Syntology's instrument failed) · 10 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 22 samples from 2 repositories, 12 ran (7 with no instrument failure; 5 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 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 2 of the 12 samples that ran have no recorded date.

zeju1997/oft mentioned on GitHubjaxMITSyntology: 19 samples, 9 ran (3 instrument-failed) · 10 unverified report
zqiu24/oft mentioned on GitHubjaxSyntology: 3 samples, 3 ran (2 instrument-failed) · 0 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

22 samples harvested: 12 ran (of which 1 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 5 where Syntology's instrument failed) · 10 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.

5ran · our draft was wrong
7ran
10unverified

Licence: 2 of the 22 samples are pointer only, meaning Syntology does not serve that copy's text. For 2 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 2 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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 1 sample 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.

zeju1997/oft community (archive-listed) 19 samples, 9 ran (3 instrument-failed) · 10 unverified
calculate_metrics in 61 papers oft-control/eval_canny.py ran MIT (permissive) · a73565d61cbda080 · report fixture: signature pred:ndarray/2/bool, target:ndarray/2/bool · returned tuple · level recorded 2026-09-24
collate_fn in 500 papers oft-db/hyperparameter_oft.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · 229d634611bdc645 · report fixture: signature examples:list/1/dict, with_prior_preservation:bool/0/bool · returned dict · level recorded, date not recorded
find_directories_missing_folders oft-db/check_empty.py ran fingerprinted MIT (permissive) · bca0e14a517d825c · report fixture: signature directory:str/0/str, folder_names:list/1/str · returned list · level recorded 2026-09-24
find_subdirectory_names_without_png oft-db/check_empty.py ran MIT (permissive) · 33aebbf50ad78324 · report fixture: signature directory:str/0/str · returned list · level recorded 2026-09-24
get_node_name in 18 papers oft-control/tool_add_control_oft.py ran · our draft was wrong MIT (permissive) · 0dbf855dab1ab575 · report fixture: signature name:str/0/str, parent_name:str/0/str · returned tuple · level recorded, date not recorded
get_params_to_save oft-db/train_dreambooth_flax.py ran MIT (permissive) · 320666d58c95245c · report fixture: signature params:dict/0/? · returned dict · level recorded 2026-09-25
get_state_dict in 105 papers oft-control/oldm/model.py ran · our draft was wrong MIT (permissive) · cfc17707f35e7ec4 · report fixture: signature d:dict/0/? · returned dict · level recorded 2026-09-25
plot_kpts oft-control/eval_landmark.py ran MIT (permissive) · 1ff5174d4258a8ab · report fixture: signature image:ndarray/3/int, kpts:ndarray/2/float, color:str/0/str · returned ndarray · level recorded 2026-09-24
plot_points oft-control/eval_landmark.py ran MIT (permissive) · 9320471cb08ee42e · report fixture: signature image:ndarray/3/int, kpts:ndarray/2/float, color:str/0/str · returned ndarray · level recorded 2026-09-24
clip_image in 6 papers oft-db/eval_ablation.py unverified MIT (permissive) · 862b8dc1d24afa80 · report
clip_text oft-db/eval_ablation.py unverified MIT (permissive) · 028c10fb01328ebd · report
dino oft-db/eval_ablation.py unverified MIT (permissive) · 8c1f9091bcfcc35f · report
get_full_repo_name in 18 papers oft-db/train_dreambooth_flax.py unverified MIT (permissive) here; the stored copy is pointer only · 6431708d8f3ff177 · report
import_model_class_from_model_name_or_path in 64 papers oft-db/hyperparameter_oft.py unverified MIT (permissive) · 2f00b39db5a81466 · report
inject_trainable_oft oft-control/oft.py unverified MIT (permissive) · 5cfa761d3c77fa2d · report
load_state_dict in 114 papers oft-control/oldm/model.py unverified MIT (permissive) · df733a879693145d · report
parse_args in 355 papers oft-db/hyperparameter_oft.py unverified MIT (permissive) · 9a1b8f02d12708b3 · report
parse_args in 355 papers oft-db/train_dreambooth.py unverified MIT (permissive) · f5132e29bee34d4d · report
parse_args in 355 papers oft-db/train_dreambooth_lora.py unverified MIT (permissive) · 2f3af25e8d9d78f5 · report
zqiu24/oft community (archive-listed) 3 samples, 3 ran (2 instrument-failed) · 0 unverified
OFTInjectedLinear oft-control/oft.py ran · metamorphic tier: deterministic MIT (permissive) · e85418d3d1fc5999 · report fixture: returned OFTInjectedLinear · level recorded 2026-09-22
project in 44 papers
census: 8 implementations from 9 papers, 7 distinct outputs across 3 buckets, one shared input per bucket
oft-control/oft.py ran · our draft was wrong fingerprinted MIT (permissive) · ccec8c1b07c1a3b6 · report fixture: signature R:Tensor/2/float, eps:float/0/float · returned Tensor · level recorded 2026-09-22
project_batch oft-control/oft.py ran · our draft was wrong fingerprinted MIT (permissive) · 4cd1d6571b9e35a1 · report fixture: signature R:Tensor/3/float, eps:float/0/float · returned Tensor · level recorded 2026-09-22

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 Orthogonal Finetuning (OFT) (graph method: Orthogonal Finetuning (OFT)). 3 bricks; individually on a fixture: 3 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

  1. OFTInjectedLinear entry · level 1 · ran · e85418d3d1fc5999
  2. project helper · level 1 · ran · our draft was wrong · ccec8c1b07c1a3b6
  3. project_batch helper · level 1 · ran · our draft was wrong · 4cd1d6571b9e35a1

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="2306.07280") returns these assemblies with each brick's role and wiring (how to connect).

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