Papers › Powering Up Zeroth-Order Training via Subspace Gradient Orthogonalization
Powering Up Zeroth-Order Training via Subspace Gradient Orthogonalization
Yicheng Lang, Changsheng Wang, Yihua Zhang, Mingyi Hong, Zheng Zhang, Wotao Yin, Sijia Liu
Title, abstract, authors and date from arXiv's metadata (CC0); this paper is not in the Papers with Code archive (frozen 2025-07-28).
Zeroth-order (ZO) optimization provides a gradient-free alternative to first-order (FO) methods by estimating gradients via finite differences of function evaluations, and has recently emerged as a memory-efficient paradigm for fine-tuning large-scale models by avoiding backpropagation. However, ZO optimization has a fundamental tension between accuracy and query efficiency. In this work, we show that ZO optimization can be substantially improved by unifying two complementary principles: (i) a projection-based subspace view that reduces gradient estimation variance by exploiting the intrinsic low-rank structure of model updates, and (ii) Muon-style spectral optimization that applies gradient orthogonalization to extract informative spectral structure from noisy ZO gradients. These findings form a unified framework of subspace gradient orthogonalization, which we instantiate in a new method, ZO-Muon, admitting a natural interpretation as a low-rank Muon optimizer in the ZO setting. Extensive experiments on large language models (LLMs) and vision transformers (ViTs) demonstrate that ZO-Muon significantly accelerates convergence and achieves a win-win improvement in accuracy and query/runtime efficiency. Notably, compared to the popular MeZO baseline, ZO-Muon requires only 24.7% of the queries to reach the same SST-2 performance for LLM fine-tuning, and improves accuracy by 25.1% on ViT-B fine-tuning on CIFAR-100.
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):
get_harvested_code_for_paper(arxiv_id="2602.17155")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="2602.17155", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="2602.17155")
Code
Syntology 15 code samples harvested from 1 repository linked to this paper: 8 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 2 honoured, 0 violated, 4 with no contract checked; 2 where Syntology's instrument failed) · 7 unverified. Where Syntology's instrument failed: 2 where our contract draft was wrong; failures of Syntology's instrument, not of the code.
By repository: found in paper text by Syntology: 15 samples from 1 repository, 8 ran (6 with no instrument failure; 2 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 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 8 samples that ran has no recorded date.
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
15 samples harvested: 8 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 2 honoured, 0 violated, 4 with no contract checked; 2 where Syntology's instrument failed) · 7 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.
Licence: 3 of the 15 samples are pointer only, meaning Syntology does not serve that copy's text. For 3 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 OPTML-Group/ZO-Muon. “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.
b989d4bce26f77ce · report
fixture: signature pred:str/0/str, gold:list/1/str · returned float64 · level recorded 2026-09-26 bd2c31732da89465 · report
fixture: signature w_shape:tuple/1/int, device:str/0/str, dtype:dtype/0/?, rank:int/0/int · returned tuple · level recorded 2026-09-26 b5adfd6ac99a1b18 · report
fixture: signature root_module:callable/0/?, key:str/0/str · returned tuple · level recorded 2026-09-22 4a22da68f93d77ed · report
fixture: signature learning_rate:float/0/float, name:str/0/str, num_warmup_steps:int/0/int, num_decay_steps:int/0/int, current_step:int/0/int, num_training_steps:int/0/int · returned float · level recorded 2026-09-26 efcead5088fac5fa · report
fixture: signature learning_rate:float/0/float, name:str/0/str, num_warmup_steps:int/0/int, num_decay_steps:int/0/int, current_step:int/0/int, num_training_steps:int/0/int · returned float · level recorded 2026-09-26 c04f796282ba4842 · report
fixture: signature learning_rate:float/0/float, name:str/0/str, num_warmup_steps:int/0/int, num_decay_steps:int/0/int, current_step:int/0/int, num_training_steps:int/0/int · returned float · level recorded 2026-09-26 e7e75981cb464788 · report
fixture: signature s:str/0/str · returned str · level recorded, date not recorded 26e94eb74c93a705 · report
fixture: signature G:Tensor/2/float · returned Tensor · level recorded 2026-09-22 84c658f2c9252878 · report
66ebbc67206f9124 · report
84920a8095dca4fa · report
635c93a80a199840 · report
d3a6a969126c5294 · report
Results from the paper
The Papers with Code archive ends with its 2025-07-28 snapshot. This paper's arXiv identifier, 2602.17155, was issued in February 2026, after that date, so the archive has no leaderboard rows for it.
Placed on leaderboards by Syntology Syntology
Syntology's extractor, the model Claude Sonnet 4.5, judged this paper's tables to report on 5 leaderboards (a model's judgement, not a result) and has not placed the paper on any of them: Natural Language Inference · RTE (refused by a rule: the table's column header does not name the leaderboard's metric); parameter-efficient fine-tuning · BoolQ (refused by a rule: the table's column header does not name the leaderboard's metric); Text Classification · GLUE RTE (refused by a rule: the table's column header does not name the leaderboard's metric); Text Classification · GLUE SST2 (refused by a rule: the table's column header does not name the leaderboard's metric); Text Classification · SST-2 (refused by a rule: the table's column header does not name the leaderboard's metric). What is not shown.
Syntology has checked 8,886 of the 18,637 papers on this site that are newer than the archive (for 3,717 of them no archive leaderboard matched the paper's tables, so there was nothing further to check); 733 were read and have nothing to place (arXiv has no HTML version of the paper, or that version has no tables), 41 could not be read (the extractor's reply could not be parsed), and results from the other 8,977 appear after they are checked.
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