SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2207.10049 · ICML · 2022

Pretraining a Neural Network before Knowing Its Architecture

Boris Knyazev

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 3 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2207.10049.json · badge: /api/ran/2207.10049.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
facebookresearch/ppuda canonical (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
yaya-yns/tart extension (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
graph_pooling Ran
ran; our contract draft was wrong, not the code
yaya-yns/tart/experiments/neural_predictor.py
code served (permissive licence) · get_code("e5789243721bd286")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# inputs: shape (batch, num_vertices_per_graph, feature_dim)
# We'll use batch=2, max_num_vertices=5, feature_dim=4
inputs = torch.randn(2, 5, 4)

# num_vertices: shape (batch,) - number of vertices for each graph in the batch
num_vertices = torch.tensor([5.0, 3.0])

ARGS = {
    "inputs": inputs,
    "num_vertices": num_vertices
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
normalize_adj Ran
ran; our contract draft was wrong, not the code
yaya-yns/tart/experiments/neural_predictor.py
code served (permissive licence) · get_code("dfa62dfebede20fc")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned Tensor, ndim 3, dtype family float, shape (2, 3, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function expects an adjacency matrix with at least 3 dimensions
# It sums over dimension 2 (the third dimension) and divides
# Let's create a 3D tensor: (batch_size, num_nodes, num_nodes)
# Using distinct sizes: batch=2, nodes1=3, nodes2=5
adj = torch.randn(2, 3, 5)

ARGS = {
    'adj': adj
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 3, 5) — shape (2, 3, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
to_cuda Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yaya-yns/tart/experiments/neural_predictor.py
code served (permissive licence) · get_code("54e37c22f817d98b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned dict. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Create a mixed object structure to test all branches of to_cuda
tensor1 = torch.randn(2, 4)
tensor2 = torch.randn(3, 5)

obj = {
    'tensor': tensor1,
    'tuple': (tensor2, torch.randn(2, 3)),
    'list': [torch.randn(1, 4), torch.randn(2, 2)],
    'nested_dict': {
        'inner_tensor': torch.randn(3, 4),
        'int_val': 42,
        'float_val': 3.14,
        'str_val': 'hello'
    },
    'int': 10,
    'float': 2.5,
    'str': 'test'
}

ARGS = {'obj': obj}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
ghn_parallel Not yet run
unverified (harvested, no recorded run)
facebookresearch/ppuda/ppuda/ghn/nn.py
code served (permissive licence) · get_code("51210ced01277bb6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 2 later) and 0 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
normalize_adj 3 other papers
2 dated earlier, 1 later
1810.05749 (2018-10)
Graph HyperNetworks for Neural Architecture Search
1912.00848 (2019-12)
Neural Predictor for Neural Architecture Search
2210.14016 (2022-10)
Shortest Edit Path Crossover: A Theory-driven Solution to the Permutation Problem in Evolutionary Neural Architecture Search
ghn_parallel 2 other papers
1 dated earlier, 1 later
2110.13100 (2021-10)
Parameter Prediction for Unseen Deep Architectures
2303.04143 (2023-03)
Can We Scale Transformers to Predict Parameters of Diverse ImageNet Models?
graph_pooling 2 other papers
2 dated earlier
1810.05749 (2018-10)
Graph HyperNetworks for Neural Architecture Search
1912.00848 (2019-12)
Neural Predictor for Neural Architecture Search
to_cuda 2 other papers
2 dated earlier
1810.05749 (2018-10)
Graph HyperNetworks for Neural Architecture Search
1912.00848 (2019-12)
Neural Predictor for Neural Architecture Search

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Training large neural networks is possible by training a smaller hypernetwork that predicts parameters for the large ones. A recently released Graph HyperNetwork (GHN) trained this way on one million smaller ImageNet architectures is able to predict parameters for large unseen networks such as ResNet-50. While networks with predicted parameters lose performance on the source task, the predicted parameters have been found useful for fine-tuning on other tasks. We study if fine-tuning based on the same GHN is still useful on novel strong architectures that were published after the GHN had been trained. We found that for recent architectures such as ConvNeXt, GHN initialization becomes less useful than for ResNet-50. One potential reason is the increased distribution shift of novel architectures from those used to train the GHN. We also found that the predicted parameters lack the diversity necessary to successfully fine-tune parameters with gradient descent. We alleviate this limitation by applying simple post-processing techniques to predicted parameters before fine-tuning them on a target task and improve fine-tuning of ResNet-50 and ConvNeXt.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2207.10049")
get_code_for_paper("2207.10049")
have("2207.10049")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2207.10049.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2207.10049.svg)](https://syntology.ai/paper/2207.10049)

Connect an agent — have() is free.