Papers › Learning to Learn with Generative Models of Neural Network Checkpoints

Learning to Learn with Generative Models of Neural Network Checkpoints

26 Sep 2022arXiv:2209.12892archive 2025-07-28

William Peebles, Ilija Radosavovic, Tim Brooks, Alexei A. Efros, Jitendra Malik

We explore a data-driven approach for learning to optimize neural networks. We construct a dataset of neural network checkpoints and train a generative model on the parameters. In particular, our model is a conditional diffusion transformer that, given an initial input parameter vector and a prompted loss, error, or return, predicts the distribution over parameter updates that achieve the desired metric. At test time, it can optimize neural networks with unseen parameters for downstream tasks in just one update. We find that our approach successfully generates parameters for a wide range of loss prompts. Moreover, it can sample multimodal parameter solutions and has favorable scaling properties. We apply our method to different neural network architectures and tasks in supervised and reinforcement learning.

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 20 code samples harvested from 1 repository linked to this paper: 16 ran (of which 0 constructed an object rather than computing a result; 15 with no instrument failure: 0 honoured, 0 violated, 15 with no contract checked; 1 where Syntology's instrument failed) · 4 unverified. Where Syntology's instrument failed: 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 20 samples from 1 repository, 16 ran (15 with no instrument failure; 1 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.

wpeebles/g.pt officialmentioned in papermentioned on GitHubpytorchBSD-2-ClauseSyntology: 20 samples, 16 ran (1 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

20 samples harvested: 16 ran (of which 0 constructed an object rather than computing a result; 15 with no instrument failure: 0 honoured, 0 violated, 15 with no contract checked; 1 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 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; the contract check could not run
15ran
4unverified

Licence: 1 of the 20 samples is pointer only, meaning Syntology does not serve that copy's text. For that sample, 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 wpeebles/g.pt. “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. 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.

wpeebles/g.pt official repository 20 samples, 16 ran (1 instrument-failed) · 4 unverified
all_gather in 119 papers Gpt/distributed.py ran BSD-2-Clause (permissive) · 250494257e1f8188 · report fixture: signature input:Tensor/2/float, cat:bool/0/bool, dim:int/0/int · returned Tensor · level recorded 2026-09-25
color_norm data_gen/train_cifar10.py ran BSD-2-Clause (permissive) · d160a205c24a70f9 · report fixture: signature im:Tensor/3/float, mean:Tensor/1/float, std:Tensor/1/float · returned Tensor · level recorded 2026-09-24
construct_loader in 6 papers data_gen/train_mnist.py ran BSD-2-Clause (permissive) · 337c2966cca20001 · report fixture: signature dataset:TensorDataset/0/?, mb_size:int/0/int, shuffle:bool/0/bool, drop_last:bool/0/bool · returned DataLoader · level recorded 2026-09-25
create_thresholding_fn Gpt/vis.py ran BSD-2-Clause (permissive) · 9759670b96772cd5 · report fixture: signature thresholding:str/0/str, param_range:list/1/float · returned function · level recorded 2026-09-25
ddim_synth playground.py ran BSD-2-Clause (permissive) · 06ebeb225cfdc4a0 · report fixture: signature diffusion:callable/0/?, G:callable/0/?, loss_target:Tensor/2/float, loss_prev:Tensor/2/float, w_prev:Tensor/3/float, ddim_sample_loop_kwargs:dict/0/? · returned Tensor · level recorded 2026-09-25
dump_json_stats Gpt/meters.py ran BSD-2-Clause (permissive) · 9817cd76b1cffb0c · report fixture: signature stats:dict/0/? · returned str · level recorded 2026-09-25
eta_str data_gen/train_mnist.py ran BSD-2-Clause (permissive) · 5c827254b86e623d · report fixture: signature eta_td:timedelta/0/? · returned str · level recorded 2026-09-25
float_to_decimal Gpt/meters.py ran BSD-2-Clause (permissive) · ed27fb884352e87a · report fixture: signature data:dict/0/?, prec:int/0/int · returned dict · level recorded 2026-09-25
horizontal_flip in 12 papers data_gen/train_cifar10.py ran fingerprinted BSD-2-Clause (permissive) · 1cd4047c59310f59 · report fixture: signature im:ndarray/3/int, p:float/0/float · returned ndarray · level recorded 2026-09-25
is_main_proc Gpt/distributed.py ran BSD-2-Clause (permissive) · 13bc2d293163e97b · report fixture: signature local:bool/0/bool · returned bool · level recorded 2026-09-25
lr_fun_cos Gpt/utils.py ran BSD-2-Clause (permissive) · 3beb27a8d68b5a5e · report fixture: signature cfg:Config/0/?, cur_epoch:int/0/int · returned float64 · level recorded 2026-09-25
scaled_all_reduce Gpt/distributed.py ran BSD-2-Clause (permissive) · a45c474751d02fe9 · report fixture: signature cfg:Config/0/?, tensors:list/1/Tensor · returned list · level recorded 2026-09-25
slerp in 28 papers
census: 6 implementations from 6 papers, 6 distinct outputs across 2 buckets, one shared input per bucket
Gpt/latent_walk_helpers.py ran BSD-2-Clause (permissive) · c2e26f19600bfc6a · report fixture: signature low:Tensor/2/float, high:Tensor/2/float, weight:Tensor/1/float · returned Tensor · level recorded 2026-09-25
slerpify Gpt/latent_walk_helpers.py ran fingerprinted BSD-2-Clause (permissive) · d06e1fd0a7f317f7 · report fixture: signature z:Tensor/3/float, n_steps:int/0/int · returned Tensor · level recorded 2026-09-25
time_string Gpt/meters.py ran fingerprinted BSD-2-Clause (permissive) · 803bcd86e3301891 · report fixture: signature seconds:float/0/float · returned str · level recorded 2026-09-24
zero_pad in 9 papers data_gen/train_cifar10.py ran; the contract check could not run (RAISES) fingerprinted BSD-2-Clause (permissive) here; the stored copy is pointer only · 8e712c383c677282 · report fixture: signature im:ndarray/3/float, pad_size:int/0/int · returned ndarray · level recorded 2026-09-22
construct_loader in 6 papers Gpt/utils.py unverified BSD-2-Clause (permissive) · 7bef924189a2a628 · report
setup_env Gpt/utils.py unverified BSD-2-Clause (permissive) · 6f53e4795c83768e · report
synth Gpt/vis.py unverified BSD-2-Clause (permissive) · 8c81a09dfbb6d408 · report
top1_error data_gen/train_mnist.py unverified BSD-2-Clause (permissive) · 093a895f2d9c1921 · report

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

DiffusionTest

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