SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1805.08079 · 2018

Faster Neural Network Training with Approximate Tensor Operations

arXiv · PDF · Open in the Atlas

Code that ran

4 of 9 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; 2 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 failures of our instrument; 5 no recorded run), measured 2026-11-21; 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/1805.08079.json · badge: /api/ran/1805.08079.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
acsl-technion/approx canonical (classifier label)
classifier basis: census
4 ran · 5 no recorded run
FunctionStatusWhere it lives
approx_linear_forward Ran
ran on a synthesized input
acsl-technion/approx/src/pytorch/approx_mul_pytorch/functional/approx_linear.py
code served (permissive licence) · get_code("100ff7853930b303")
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 Tensor, ndim 2, dtype family float, shape (2, 3) 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
import torch.nn as nn

torch.manual_seed(0)

# Define tensor dimensions with distinct sizes:
# - batch_size: 2
# - in_features: 4
# - out_features: 3

batch_size = 2
in_features = 4
out_features = 3

# Create input tensor: (batch_size, in_features)
input_tensor = torch.randn(batch_size, in_features)

# Create weight tensor: (out_features, in_features)
# This will be transposed in approx_linear_forward, so shape is correct for xA^T
weight_tensor = torch.randn(out_features, in_features)

# Create bias tensor: (out_features,)
bias_tensor = torch.randn(out_features)

# Define sampling parameters
sample_ratio = 0.5  # sample 50% of input features
minimal_k = 1       # minimum number of features to keep
sample_ratio_bwd = 0.5  # backward pass sampling ratio
minimal_k_bwd = 1   # minimum k for backward
sample_ratio_wu = 0.5   # weight update sampling ratio
minimal_k_wu = 1    # minimum k for weight update

ARGS = {
    'input': input_tensor,
    'weight': weight_tensor,
    'bias': bias_tensor,
    'sample_ratio': sample_ratio,
    'minimal_k': minimal_k,
    'sample_ratio_bwd': sample_ratio_bwd,
    'minimal_k_bwd': minimal_k_bwd,
    'sample_ratio_wu': sample_ratio_wu,
    'minimal_k_wu': minimal_k_wu,
}
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 Tensor, ndim 2, dtype family float, shape (2, 3) — shape (2, 3) 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
approx_linear_forward_xA_b Ran
ran on a synthesized input
acsl-technion/approx/src/pytorch/approx_mul_pytorch/functional/approx_linear.py
code served (permissive licence) · get_code("2584e5b153d87f30")
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 Tensor, ndim 2, dtype family float, shape (2, 3) 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
import torch.nn as nn

torch.manual_seed(0)

# Create small tensors with distinct dimensions
batch_size = 2
in_features = 4
out_features = 3

# input: (batch_size, in_features)
input_tensor = torch.randn(batch_size, in_features)

# weight: (in_features, out_features)
weight_tensor = torch.randn(in_features, out_features)

# bias: (out_features,)
bias_tensor = torch.randn(out_features)

# Hyperparameters
sample_ratio = 0.5
minimal_k = 1
sample_ratio_bwd = 0.5
minimal_k_bwd = 1
sample_ratio_wu = 0.5
minimal_k_wu = 1

ARGS = {
    'input': input_tensor,
    'weight': weight_tensor,
    'bias': bias_tensor,
    'sample_ratio': sample_ratio,
    'minimal_k': minimal_k,
    'sample_ratio_bwd': sample_ratio_bwd,
    'minimal_k_bwd': minimal_k_bwd,
    'sample_ratio_wu': sample_ratio_wu,
    'minimal_k_wu': minimal_k_wu,
}
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 Tensor, ndim 2, dtype family float, shape (2, 3) — shape (2, 3) 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
get_hms Ran
ran, honoured the contract we drafted
acsl-technion/approx/src/pytorch/cifar_dist/config.py
code served (permissive licence) · get_code("bb104479d26e39c3")
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) and returned tuple, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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)

ARGS = {
    "seconds": 3661
}
Setup block drafted by
a language model, Claude Haiku 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 tuple, ndim 1, dtype family int — a tuple output has no array shape of its own; none is stated
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
learning_rate Ran
ran, honoured the contract we drafted
acsl-technion/approx/src/pytorch/cifar_dist/config.py
code served (permissive licence) · get_code("947a620765d795b2")
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 float, ndim 0, dtype family float, 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 math

ARGS = {
    "init": 0.1,
    "epoch": 100
}
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 float, ndim 0, dtype family float, 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
conv2d_bwd_bernoulli_sampling Not yet run
unverified (harvested, no recorded run)
acsl-technion/approx/src/pytorch/approx_mul_pytorch/functional/approx_conv2d.py
code served (permissive licence) · get_code("3f584bce253f07ce")
How it ran

no invocation recorded.

conv2d_bwd_random_sampling Not yet run
unverified (harvested, no recorded run)
acsl-technion/approx/src/pytorch/approx_mul_pytorch/functional/approx_conv2d.py
code served (permissive licence) · get_code("21c5ffd3a309185b")
How it ran

no invocation recorded.

conv2d_bwd_topk Not yet run
unverified (harvested, no recorded run)
acsl-technion/approx/src/pytorch/approx_mul_pytorch/functional/approx_conv2d.py
code served (permissive licence) · get_code("bee59f4bc71ec988")
How it ran

no invocation recorded.

get_mnist Not yet run
unverified (harvested, no recorded run)
acsl-technion/approx/src/pytorch/mnist_mlp_pytorch/data.py
code served (permissive licence) · get_code("485595ea85ffbcc1")
How it ran

no invocation recorded.

linear_top_k Not yet run
unverified (harvested, no recorded run)
acsl-technion/approx/src/pytorch/approx_mul_pytorch/functional/approx_linear.py
code served (permissive licence) · get_code("4a2121e58100f403")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 2 later) and 7 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
get_hms 7 other papers
1 dated earlier, 6 later
1802.07088 (2018-02)
i-RevNet: Deep Invertible Networks
1812.05159 (2018-12)
An Empirical Study of Example Forgetting during Deep Neural Network Learning
1905.10964 (2019-05)
Combating Label Noise in Deep Learning Using Abstention
and 4 more not named here
learning_rate 5 other papers
1 dated earlier, 4 later
1802.07088 (2018-02)
i-RevNet: Deep Invertible Networks
1904.05514 (2019-04)
Mitigating Information Leakage in Image Representations: A Maximum Entropy Approach
2204.10839 (2022-04)
How Sampling Impacts the Robustness of Stochastic Neural Networks
and 2 more not named here

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 2026-11-21.

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 propose a novel technique for faster deep neural network training which systematically applies sample-based approximation to the constituent tensor operations, i.e., matrix multiplications and convolutions. We introduce new sampling techniques, study their theoretical properties, and prove that they provide the same convergence guarantees when applied to SGD training. We apply approximate tensor operations to single and multi-node training of MLP and CNN networks on MNIST, CIFAR-10 and ImageNet datasets. We demonstrate up to 66% reduction in the amount of computations and communication, and up to 1.37x faster training time while maintaining negligible or no impact on the final test accuracy.

For agents

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

get_harvested_code_for_paper("1805.08079")
get_code_for_paper("1805.08079")
have("1805.08079")

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

curl https://syntology.ai/api/ran/1805.08079.json

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

[![Syntology run record](https://syntology.ai/api/ran/1805.08079.svg)](https://syntology.ai/paper/1805.08079)

Connect an agent — have() is free.