Papers › Mixture-of-Subspaces in Low-Rank Adaptation

Mixture-of-Subspaces in Low-Rank Adaptation

16 Jun 2024arXiv:2406.11909archive 2025-07-28

Taiqiang Wu, Jiahao Wang, Zhe Zhao, Ngai Wong

In this paper, we introduce a subspace-inspired Low-Rank Adaptation (LoRA) method, which is computationally efficient, easy to implement, and readily applicable to large language, multimodal, and diffusion models. Initially, we equivalently decompose the weights of LoRA into two subspaces, and find that simply mixing them can enhance performance. To study such a phenomenon, we revisit it through a fine-grained subspace lens, showing that such modification is equivalent to employing a fixed mixer to fuse the subspaces. To be more flexible, we jointly learn the mixer with the original LoRA weights, and term the method Mixture-of-Subspaces LoRA (MoSLoRA). MoSLoRA consistently outperforms LoRA on tasks in different modalities, including commonsense reasoning, visual instruction tuning, and subject-driven text-to-image generation, demonstrating its effectiveness and robustness. Codes are available at https://github.com/wutaiqiang/MoSLoRA.

PaperPDFCodeCode 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="2406.11909")

Code

Syntology Ran 4 of 6 code samples harvested from 1 repository linked to this paper; 2 have no recorded run. Of those that ran: 3 ran · our draft was wrong; 1 ran with no contract checked.

By repository: official repository: 6 samples from 1 repository, 4 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

wutaiqiang/moslora officialmentioned in papermentioned on GitHubpytorch 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

6 samples harvested; 4 ran; 0 honoured the contract we drafted; 2 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.

3ran · our draft was wrong
1ran
2unverified

Licence: 6 of the 6 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 wutaiqiang/moslora. “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.

determine_scheduler_type wutaiqiang/moslora/subject_driven_generation/train_dreambooth_lora_sdxl.py official repository ran no licence file found · pointer only · 03e1c6795642a8db · report
generate_prompt wutaiqiang/moslora/commonsense_reasoning/finetune.py official repository ran · our draft was wrong no licence file found · pointer only · 56e674752db00649 · report
generate_prompt wutaiqiang/moslora/commonsense_reasoning/commonsense_evaluate.py official repository ran · our draft was wrong fingerprinted no licence file found · pointer only · 1c6b8be890b6118a · report
load_data wutaiqiang/moslora/commonsense_reasoning/commonsense_evaluate.py official repository ran · our draft was wrong no licence file found · pointer only · 56a9466ff35b35bb · report
import_model_class_from_model_name_or_path wutaiqiang/moslora/subject_driven_generation/train_dreambooth_lora_sdxl.py official repository unverified no licence file found · pointer only · 86965b6bae051ea4 · report
main wutaiqiang/moslora/commonsense_reasoning/commonsense_evaluate.py official repository unverified no licence file found · pointer only · b8da7d64c2442014 · report

Tasks

Common Sense ReasoningImage GenerationQuestion AnsweringSentence CompletionText to Image GenerationText-to-Image GenerationVisual Question Answering

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Common Sense Reasoning ARC (Challenge) LLaMA 3 8B + MoSLoRA (fine-tuned) Accuracy 81.5 #16 of 54 Archive leaderboard report
Common Sense Reasoning ARC (Easy) LLaMA 3 8B+MoSLoRA (fine-tuned) Accuracy 90.5 #2 of 47 Archive leaderboard report
Common Sense Reasoning WinoGrande LLaMA3 8B+MoSLoRA Accuracy 85.8 #10 of 77 Archive leaderboard report
Question Answering BoolQ LLaMA3+MoSLoRA Accuracy 74.6 #38 of 65 Archive leaderboard report
Question Answering OpenBookQA LLaMA-3 8B+MoSLoRA Accuracy 86.8 #14 of 45 Archive leaderboard report
Question Answering PIQA LLaMA3 8B+MoSLoRA Accuracy 89.7 #2 of 67 Archive leaderboard report
Question Answering SIQA LLaMA-3 8B+MoSLoRA (fine-tuned) Accuracy 81.0 #5 of 24 Archive leaderboard report
Sentence Completion HellaSwag LLaMA3+MoSLoRA Accuracy 95.0 #5 of 89 Archive leaderboard report
Visual Question Answering MM-Vet LLaVA-InternLM2-7B-ViT + MoSLoRA GPT-4 score 35.2 #159 of 231 Archive leaderboard report
Visual Question Answering MM-Vet InternLM2+ViT (QMoSLoRA) GPT-4 score 35.2 #160 of 231 Archive leaderboard report
Visual Question Answering MMBench LLaVA-InternLM2-ViT + MoSLoRA GPT-3.5 score 73.8 #1 of 5 Archive leaderboard report
Visual Question Answering MMBench LLaVA-LLaMA3-8B-ViT + MoSLoRA GPT-3.5 score 73.0 #3 of 5 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

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