Papers › DARE the Extreme: Revisiting Delta-Parameter Pruning For Fine-Tuned Models

DARE the Extreme: Revisiting Delta-Parameter Pruning For Fine-Tuned Models

12 Oct 2024arXiv:2410.09344archive 2025-07-28

Wenlong Deng, Yize Zhao, Vala Vakilian, Minghui Chen, Xiaoxiao Li, Christos Thrampoulidis

Storing open-source fine-tuned models separately introduces redundancy and increases response times in applications utilizing multiple models. Delta-parameter pruning (DPP), particularly the random drop and rescale (DARE) method proposed by Yu et al., addresses this by pruning the majority of delta parameters--the differences between fine-tuned and pre-trained model weights--while typically maintaining minimal performance loss. However, DARE fails when either the pruning rate or the magnitude of the delta parameters is large. We highlight two key reasons for this failure: (1) an excessively large rescaling factor as pruning rates increase, and (2) high mean and variance in the delta parameters. To push DARE's limits, we introduce DAREx (DARE the eXtreme), which features two algorithmic improvements: (1) DAREx-q, a rescaling factor modification that significantly boosts performance at high pruning rates (e.g., >30 % on COLA and SST2 for encoder models, with even greater gains in decoder models), and (2) DAREx-L2, which combines DARE with AdamR, an in-training method that applies appropriate delta regularization before DPP. We also demonstrate that DAREx-q can be seamlessly combined with vanilla parameter-efficient fine-tuning techniques like LoRA and can facilitate structural DPP. Additionally, we revisit the application of importance-based pruning techniques within DPP, demonstrating that they outperform random-based methods when delta parameters are large. Through this comprehensive study, we develop a pipeline for selecting the most appropriate DPP method under various practical scenarios.

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 17 code samples harvested from 1 repository linked to this paper: 13 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 2 violated, 8 with no contract checked; 3 where Syntology's instrument failed) · 4 unverified. Where Syntology's instrument failed: 2 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: 17 samples from 1 repository, 13 ran (10 with no instrument failure; 3 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 2026-09-24 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 2 of the 13 samples that ran have no recorded date.

vengdeng/darex officialmentioned in papermentioned on GitHubpytorchSyntology: 17 samples, 13 ran (3 instrument-failed) · 4 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

17 samples harvested: 13 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 2 violated, 8 with no contract checked; 3 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.

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

Licence: 17 of the 17 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 vengdeng/darex. “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.

vengdeng/darex official repository 17 samples, 13 ran (3 instrument-failed) · 4 unverified
is_number in 70 papers utils/evaluate_llms_utils.py ran · violated contract no licence file found · pointer only · 0fe9cf6c00ef56ea · report fixture: signature s:str/0/str · returned bool · level recorded, date not recorded
add_back_decoder utils/deltaprune.py ran no licence file found · pointer only · f610ce38c215e0b3 · report fixture: signature pretrain:callable/0/?, fine_model:callable/0/?, delta_dict:dict/0/? · returned dict · level recorded 2026-09-24
batch_data in 20 papers utils/evaluate_llms_utils.py ran · our draft was wrong fingerprinted no licence file found · pointer only · d49412b3a690206e · report fixture: signature data_list:list/1/Tensor, batch_size:int/0/int · returned list · level recorded 2026-09-25
batch_data in 20 papers find_q_decoder.py ran fingerprinted no licence file found · pointer only · d49a90d96daf66f7 · report fixture: signature data_list:list/1/Tensor, batch_size:int/0/int · returned list · level recorded 2026-09-24
cal_delta_param_decoder utils/deltaprune.py ran no licence file found · pointer only · e3f5f65ef9c76959 · report fixture: signature finetune:dict/0/?, base:dict/0/? · returned dict · level recorded 2026-09-24
drop_decoder utils/deltaprune.py ran no licence file found · pointer only · d877155389b89d32 · report fixture: signature delta_param_or:dict/0/?, mask_rate:float/0/float · returned dict · level recorded 2026-09-24
extract_res utils/metrics.py ran no licence file found · pointer only · 95693ad4a192c071 · report fixture: signature test_metrics:dict/0/?, dataset_name:str/0/str · returned float · level recorded 2026-09-24
find_layers in 68 papers utils/getwx.py ran · our draft was wrong no licence file found · pointer only · 2ed16fcdc14ad951 · report fixture: signature module:callable/0/?, layers:list/1/callable, name:str/0/str · returned dict · level recorded, date not recorded
get_outpuchage utils/outputchange.py ran no licence file found · pointer only · aa111ae3259d74f6 · report fixture: signature inputs:dict/0/?, args:MinimalArgs/0/? · returned Tensor · level recorded 2026-09-24
optimal_q_calculation utils/analyticalq.py ran no licence file found · pointer only · 23f6861e67c10b94 · report fixture: signature metric_score:dict/0/?, delta_param:dict/0/?, eta:float/0/float, p:float/0/float, gpu:str/0/str, stop_iter:float/0/float · returned dict · level recorded 2026-09-24
output_loader utils/outputchange.py ran no licence file found · pointer only · 6ea7104a43c76889 · report fixture: signature seqlen:int/0/int, nsamples:int/0/int, trainenc:MockTokenizedData/0/?, tar_flag:bool/0/bool · returned list · level recorded 2026-09-24
phi_func utils/analyticalq.py ran · violated contract fingerprinted no licence file found · pointer only · 87cb5010447732ab · report fixture: signature x:ndarray/1/float · returned ndarray · level recorded 2026-09-24
remove_one_outlier find_q_encoder.py ran; the contract check could not run (RAISES) fingerprinted no licence file found · pointer only · db6d55cd1fb1ce20 · report fixture: signature data:list/1/float · returned list · level recorded 2026-09-24
analytical_resolve_new utils/analyticalq.py unverified no licence file found · pointer only · 080b98d73a404601 · report
get_loaders in 51 papers utils/getwx.py unverified no licence file found · pointer only · a54651cc8454cb77 · report
get_outpuchage_decoder utils/outputchange.py unverified no licence file found · pointer only · 6839d02854b154b2 · report
get_wikitext2 in 78 papers utils/getwx.py unverified no licence file found · pointer only · ecc04d6edadf3e8b · report

Tasks

parameter-efficient fine-tuning

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

COLAPruning

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