SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.05180 · CVPR · 2022

Knowledge Distillation as Efficient Pre-training: Faster Convergence, Higher Data-efficiency, and Better Transferability

Xiaojuan Qi, Song Bai, Jihan Yang, Ruifei He, Shuyang Sun

arXiv · PDF · Open in the Atlas

Code that ran

9 of 13 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; 6 failures of our instrument; 4 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/2203.05180.json · badge: /api/ran/2203.05180.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
CVMI-Lab/KDEP canonical (classifier label)
classifier basis not recorded on the link
9 ran · 4 no recorded run
FunctionStatusWhere it lives
conv3x3 Ran
ran; our contract draft was wrong, not the code
CVMI-Lab/KDEP/src/resnet.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from lixiaotong97/DSU, is pointer only (licence: NONE) · get_code("fac5364e2f53c6db")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from lixiaotong97/DSU, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

conv_1x1_bn Ran
ran; our contract draft was wrong, not the code
CVMI-Lab/KDEP/src/mobilenet.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from zhoudaquan/rethinking_bottleneck_design, is pointer only (licence: NOASSERTION) · get_code("a0131fb70c267a9e")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from zhoudaquan/rethinking_bottleneck_design, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

conv_1x1_bn_pre_relu Ran
ran on a synthesized input
CVMI-Lab/KDEP/src/mobilenet.py
code served (permissive licence) · get_code("82a7c55ebdb5ddd1")
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 Sequential. 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "inp": 3,
    "oup": 16,
}
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 Sequential — a Sequential output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Sequential output the recorded ndim 1 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
conv_3x3_bn Ran
ran; our contract draft was wrong, not the code
CVMI-Lab/KDEP/src/mobilenet.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from zhoudaquan/rethinking_bottleneck_design, is pointer only (licence: NOASSERTION) · get_code("408e747e0425a594")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from zhoudaquan/rethinking_bottleneck_design, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

drop_connect Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
CVMI-Lab/KDEP/src/efficientnet_utils.py
code served (permissive licence) · get_code("d3319e3d34ca90ca")
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 4, dtype family float, shape (2, 4, 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)

# Create a small input tensor with shape BCWH (Batch, Channels, Width, Height)
# Using batch_size=2, channels=4, width=3, height=5 for distinct dimensions
inputs = torch.randn(2, 4, 3, 5, dtype=torch.float32)

# Probability of drop connection
p = 0.5

# Training mode flag
training = True

ARGS = {
    "inputs": inputs,
    "p": p,
    "training": training
}
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 4, dtype family float, shape (2, 4, 3, 5) — shape (2, 4, 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
resnet18 Ran
ran on a synthesized input
CVMI-Lab/KDEP/src/resnet.py
code served (permissive licence) · get_code("bf71ee950376d6a2")
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 ResNet. 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'pretrained': False,
    'initpath': None,
}
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 ResNet — a ResNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ResNet 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
resnet18_feat Ran
ran on a synthesized input
CVMI-Lab/KDEP/src/resnet.py
code served (permissive licence) · get_code("bcbea8996447cc16")
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 ResNet_feat. 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'pretrained': False,
    'initpath': None,
}
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 ResNet_feat — a ResNet_feat output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ResNet_feat 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
round_filters Ran
ran; our contract draft was wrong, not the code
CVMI-Lab/KDEP/src/efficientnet_utils.py
code served (permissive licence) · get_code("f15a49337e69e937")
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.

from collections import namedtuple

# Create a namedtuple for global_params with the required attributes
GlobalParams = namedtuple('GlobalParams', ['width_coefficient', 'depth_divisor', 'min_depth'])

# Create valid global_params instance
global_params = GlobalParams(
    width_coefficient=1.0,
    depth_divisor=8,
    min_depth=16
)

ARGS = {
    'filters': 32,
    'global_params': global_params
}
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 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
round_repeats Ran
ran; our contract draft was wrong, not the code
CVMI-Lab/KDEP/src/efficientnet_utils.py
code served (permissive licence) · get_code("dbc0ca08d119a5a0")
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.

from collections import namedtuple

# Create a namedtuple for global_params with the depth_coefficient attribute
GlobalParams = namedtuple('GlobalParams', ['depth_coefficient'])

# Define ARGS with valid arguments for round_repeats
ARGS = {
    'repeats': 4,
    'global_params': GlobalParams(depth_coefficient=1.5)
}
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 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
RN50 Not yet run
unverified (harvested, no recorded run)
CVMI-Lab/KDEP/src/clip_prev.py
code served (permissive licence) · get_code("1dc5699e5669a244")
How it ran

no invocation recorded.

build_model Not yet run
unverified (harvested, no recorded run)
CVMI-Lab/KDEP/src/clip_prev.py
code served (permissive licence) · get_code("1d43d3fb2df8185e")
How it ran

no invocation recorded.

save_emb Not yet run
unverified (harvested, no recorded run)
CVMI-Lab/KDEP/src/entropy.py
code served (permissive licence) · get_code("6df80a5b66c15202")
How it ran

no invocation recorded.

vis_emb Not yet run
unverified (harvested, no recorded run)
CVMI-Lab/KDEP/src/entropy.py
code served (permissive licence) · get_code("e554dee428d119bb")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 distinct functions with an identity, 6 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 2 the same month, 6 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
conv3x3 686 other papers
374 dated earlier, 11 the same month, 290 later, 11 same year, order not determinable · 31 only as an assembly brick
1406.2661 (2014-06)
Generative Adversarial Networks
1409.0575 (2014-09)
ImageNet Large Scale Visual Recognition Challenge
1412.6572 (2014-12)
Explaining and Harnessing Adversarial Examples
and 683 more not named here
conv_1x1_bn 31 other papers
19 dated earlier, 1 the same month, 10 later, 1 same year, order not determinable · 2 only as an assembly brick
1703.07402 (2017-03)
Simple Online and Realtime Tracking with a Deep Association Metric
1801.04381 (2018-01)
MobileNetV2: Inverted Residuals and Linear Bottlenecks
1811.08383 (2018-11)
TSM: Temporal Shift Module for Efficient Video Understanding
and 28 more not named here
round_filters 31 other papers
20 dated earlier, 11 later · 3 only as an assembly brick
1808.08127 (2018-08)
Recalibrating Fully Convolutional Networks with Spatial and Channel 'Squeeze & Excitation' Blocks
1905.11946 (2019-05)
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
1911.09070 (2019-11)
EfficientDet: Scalable and Efficient Object Detection
and 28 more not named here
round_repeats 30 other papers
21 dated earlier, 9 later · 3 only as an assembly brick
1808.08127 (2018-08)
Recalibrating Fully Convolutional Networks with Spatial and Channel 'Squeeze & Excitation' Blocks
1905.11946 (2019-05)
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
1911.09070 (2019-11)
EfficientDet: Scalable and Efficient Object Detection
and 27 more not named here
drop_connect 23 other papers
8 dated earlier, 15 later · 2 only as an assembly brick
1905.11946 (2019-05)
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
1912.01588 (2019-12)
Leveraging Procedural Generation to Benchmark Reinforcement Learning
2010.00475 (2020-10)
FSD50K: An Open Dataset of Human-Labeled Sound Events
and 20 more not named here
conv_3x3_bn 12 other papers
5 dated earlier, 7 later
1801.04381 (2018-01)
MobileNetV2: Inverted Residuals and Linear Bottlenecks
1904.07399 (2019-04)
Adaptive Wing Loss for Robust Face Alignment via Heatmap Regression
1911.08114 (2019-11)
Neural Network Pruning with Residual-Connections and Limited-Data
and 9 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 2026-11-22.

Repositories linked to this paper

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

Abstract

Large-scale pre-training has been proven to be crucial for various computer vision tasks. However, with the increase of pre-training data amount, model architecture amount, and the private/inaccessible data, it is not very efficient or possible to pre-train all the model architectures on large-scale datasets. In this work, we investigate an alternative strategy for pre-training, namely Knowledge Distillation as Efficient Pre-training (KDEP), aiming to efficiently transfer the learned feature representation from existing pre-trained models to new student models for future downstream tasks. We observe that existing Knowledge Distillation (KD) methods are unsuitable towards pre-training since they normally distill the logits that are going to be discarded when transferred to downstream tasks. To resolve this problem, we propose a feature-based KD method with non-parametric feature dimension aligning. Notably, our method performs comparably with supervised pre-training counterparts in 3 downstream tasks and 9 downstream datasets requiring 10× less data and 5× less pre-training time. Code is available at https: //github.com/CVMI-Lab/KDEP.

For agents

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

get_harvested_code_for_paper("2203.05180")
get_code_for_paper("2203.05180")
have("2203.05180")

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

curl https://syntology.ai/api/ran/2203.05180.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.05180.svg)](https://syntology.ai/paper/2203.05180)

Connect an agent — have() is free.