SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2107.11442 · NeurIPS · 2021

Compressing Neural Networks: Towards Determining the Optimal Layer-wise Decomposition

Alaa Maalouf, Dan Feldman, Daniela Rus, Lucas Liebenwein, Oren Gal

arXiv · PDF · Open in the Atlas

Code that ran

3 of 12 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 3 have no contract verdict; 0 failures of our instrument; 9 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/2107.11442.json · badge: /api/ran/2107.11442.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
lucaslie/provable_pruning canonical (classifier label)
classifier basis: census
3 ran · 6 no recorded run
lucaslie/torchprune canonical (classifier label)
classifier basis: census
0 ran · 3 no recorded run
FunctionStatusWhere it lives
get_fig_name Ran
ran on a synthesized input
lucaslie/provable_pruning/paper/alds/script/retrain_sweep.py
code served (permissive licence) · get_code("e7aaec08393471a8")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was 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 re

def get_fig_name(title, tag):
    """Get the name of the figure with the title and tag."""
    fig_name = '_'.join(re.split('/|-|_|,', title)).replace(' ', '')
    return f'{fig_name}_sweep_{tag}.pdf'


ARGS = {
    'title': 'Accuracy/Loss-Comparison_Results,v2',
    'tag': 'experiment_001'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
key_files Ran
ran on a synthesized input
lucaslie/provable_pruning/paper/alds/script/results_viewer.py
code served (permissive licence) · get_code("4ba3c7ec40d48b25")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned int, ndim 0, dtype family int, shape () 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

def key_files(item):
    order = ['resnet20', 'resnet56', 'resnet110', 'vgg16', 'densenet22', 'wrn16_8', 'resnet18', 'resnet101', 'wide_resnet50_2', 'mobilenet_v2', 'deeplabv3_resnet50']
    for i, net in enumerate(order):
        if net in item:
            return i
    return len(order)

# The function takes a string item and checks if any model name from the order list is contained in it
# It returns the index of the first matching model name, or len(order) if no match is found

ARGS = {
    'item': 'resnet20_weights.pth'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, shape () — shape () 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
key_files Ran
ran on a synthesized input
lucaslie/provable_pruning/paper/alds/script/retrain_sweep.py
code served (permissive licence) · get_code("ef6a9bde1b92ea8d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned int, ndim 0, dtype family int, shape () 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

def key_files(item):
    order = ['resnet20', 'resnet56', 'resnet110', 'vgg16_bn', 'densenet22', 'wrn16_8', 'resnet18', 'resnet101', 'wide_resnet50_2', 'deeplabv3_resnet50']
    for i, net in enumerate(order):
        if net in item:
            return i
    return len(order)

ARGS = {
    'item': 'resnet20_checkpoint.pth'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, shape () — shape () 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
flatten_all_but_last Not yet run
unverified (harvested, no recorded run)
lucaslie/torchprune/src/torchprune/torchprune/util/tensor.py
code served (permissive licence) · get_code("cd21fa496ee07f67")
How it ran

no invocation recorded.

gather_images Not yet run
unverified (harvested, no recorded run)
lucaslie/provable_pruning/paper/lost/script/eps_ball.py
code served (permissive licence) · get_code("ef4646720ee554c1")
How it ran

no invocation recorded.

generate_noisy_images Not yet run
unverified (harvested, no recorded run)
lucaslie/provable_pruning/paper/lost/script/eps_ball.py
code served (permissive licence) · get_code("5aa7e295104184ca")
How it ran

no invocation recorded.

get_fig_name Not yet run
unverified (harvested, no recorded run)
lucaslie/provable_pruning/paper/lost/script/generalization.py
code served (permissive licence) · get_code("32d86e992080431c")
How it ran

no invocation recorded.

get_mesh Not yet run
unverified (harvested, no recorded run)
lucaslie/provable_pruning/paper/node/script/plots2d.py
code served (permissive licence) · get_code("8c4b0db81d1e76cd")
How it ran

no invocation recorded.

get_slice Not yet run
unverified (harvested, no recorded run)
lucaslie/torchprune/src/torchprune/torchprune/util/tensor.py
code served (permissive licence) · get_code("fec5d06c6e4059e2")
How it ran

no invocation recorded.

make_f_hat Not yet run
unverified (harvested, no recorded run)
lucaslie/provable_pruning/paper/lost/script/eps_ball.py
code served (permissive licence) · get_code("8eec11df7c455a60")
How it ran

no invocation recorded.

prepare_data Not yet run
unverified (harvested, no recorded run)
lucaslie/provable_pruning/paper/node/script/plots2d.py
code served (permissive licence) · get_code("7257316eff0efa77")
How it ran

no invocation recorded.

to Not yet run
unverified (harvested, no recorded run)
lucaslie/torchprune/src/torchprune/torchprune/util/tensor.py
code served (permissive licence) · get_code("e32720b53dfa1acf")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 12 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.

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

We present a novel global compression framework for deep neural networks that automatically analyzes each layer to identify the optimal per-layer compression ratio, while simultaneously achieving the desired overall compression. Our algorithm hinges on the idea of compressing each convolutional (or fully-connected) layer by slicing its channels into multiple groups and decomposing each group via low-rank decomposition. At the core of our algorithm is the derivation of layer-wise error bounds from the Eckart-Young-Mirsky theorem. We then leverage these bounds to frame the compression problem as an optimization problem where we wish to minimize the maximum compression error across layers and propose an efficient algorithm towards a solution. Our experiments indicate that our method outperforms existing low-rank compression approaches across a wide range of networks and data sets. We believe that our results open up new avenues for future research into the global performance-size trade-offs of modern neural networks.

For agents

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

get_harvested_code_for_paper("2107.11442")
get_code_for_paper("2107.11442")
have("2107.11442")

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

curl https://syntology.ai/api/ran/2107.11442.json

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

[![Syntology run record](https://syntology.ai/api/ran/2107.11442.svg)](https://syntology.ai/paper/2107.11442)

Connect an agent — have() is free.