SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2001.00705 · AAAI · 2020

Fractional Skipping: Towards Finer-Grained Dynamic CNN Inference

Zhangyang Wang, Pengfei Xu, Yingyan Lin, Yonggan Fu, Yue Wang, Jianghao Shen

arXiv · PDF · Open in the Atlas

Code that ran

16 of 24 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (3 of them constructed an object rather than computing a result; 4 honoured the contract we drafted, 0 violated it and 8 have no contract verdict; 4 failures of our instrument; 8 no recorded run), measured 2026-11-22; 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/2001.00705.json · badge: /api/ran/2001.00705.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
mit-han-lab/haq-release canonical (classifier label)
classifier basis: census
10 ran · 2 no recorded run
Torment123/DFS marked official in Papers with Code
Papers with Code archive
6 ran · 6 no recorded run
FunctionStatusWhere it lives
QuantMeasure Ran
ran on a synthesized input · constructed an object; no result was computed
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("b61afbb33cf81a80")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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.

RNNGate Ran
ran on a synthesized input · constructed an object; no result was computed
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("b6077b00a86467bb")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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.

SoftRNNGate Ran
ran on a synthesized input · constructed an object; no result was computed
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("cbdc3eb53675a6f2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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.

conv2d_biprec Ran
ran; our contract draft was wrong, not the code
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("bc6ca135467cad70")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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
mit-han-lab/haq-release/models/mobilenetv3.py
pointer only (licence: MIT); the stored copy get_code serves, from chris-boson/fashion_mnist, is pointer only (licence: NONE) · get_code("db82a4013a44fb42")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from chris-boson/fashion_mnist, 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 on a synthesized input
mit-han-lab/haq-release/models/mobilenetv2.py
code served (permissive licence) · get_code("bc660f7845db5634")
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)

# Define a custom conv layer with half_wave parameter for testing
class CustomConv2d(nn.Module):
    def __init__(self, inp, oup, kernel_size, stride, padding, bias=False, half_wave=True):
        super().__init__()
        self.conv = nn.Conv2d(inp, oup, kernel_size, stride, padding, bias=bias)
        self.half_wave = half_wave
    
    def forward(self, x):
        return self.conv(x)

# Arguments for conv_1x1_bn function
ARGS = {
    "inp": 4,
    "oup": 3,
    "conv_layer": nn.Conv2d,
    "half_wave": True
}
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_bn Ran
ran; our contract draft was wrong, not the code
mit-han-lab/haq-release/models/mobilenetv3.py
pointer only (licence: MIT); the stored copy get_code serves, from chris-boson/fashion_mnist, is pointer only (licence: NONE) · get_code("84a4cf301e929085")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from chris-boson/fashion_mnist, 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_bn Ran
ran on a synthesized input
mit-han-lab/haq-release/models/mobilenet.py
code served (permissive licence) · get_code("247c2f31dd34327d")
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)

# Define a custom conv layer that accepts half_wave parameter
class CustomConv2d(nn.Module):
    def __init__(self, inp, oup, kernel_size, stride, padding, bias=False, half_wave=True):
        super().__init__()
        self.conv = nn.Conv2d(inp, oup, kernel_size, stride, padding, bias=bias)
        self.half_wave = half_wave
    
    def forward(self, x):
        return self.conv(x)

ARGS = {
    "inp": 3,
    "oup": 16,
    "stride": 2,
    "conv_layer": nn.Conv2d,
    "half_wave": True
}
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_bn Ran
ran on a synthesized input
mit-han-lab/haq-release/models/mobilenetv2.py
code served (permissive licence) · get_code("4e64cf99f91fea92")
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)

# Create a custom conv layer that accepts half_wave parameter
class CustomConv2d(nn.Module):
    def __init__(self, in_channels, out_channels, kernel_size, stride, padding, bias=False, half_wave=True):
        super().__init__()
        self.conv = nn.Conv2d(in_channels, out_channels, kernel_size, stride, padding, bias=bias)
        self.half_wave = half_wave
    
    def forward(self, x):
        return self.conv(x)

# Test with standard nn.Conv2d
ARGS = {
    'inp': 3,
    'oup': 16,
    'stride': 1,
    'conv_layer': nn.Conv2d,
    'half_wave': True
}
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_dw Ran
ran on a synthesized input
mit-han-lab/haq-release/models/mobilenet.py
code served (permissive licence) · get_code("a972ad175e8bda7f")
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': 16,
    'oup': 32,
    'stride': 1,
    'conv_layer': nn.Conv2d
}
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
get_split_train_dataset Ran
ran on a synthesized input
mit-han-lab/haq-release/lib/utils/data_utils.py
code served (permissive licence) · get_code("8e7cf364b572ce56")
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 tuple, ndim 1, dtype family float; 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 os
import tempfile
import numpy as np
import torch
import torch.utils.data
import torchvision.transforms as transforms
from PIL import Image

# Create a temporary directory structure for imagenet
temp_dir = tempfile.mkdtemp()
data_root = os.path.join(temp_dir, 'data', 'imagenet')
traindir = os.path.join(data_root, 'train')
valdir = os.path.join(data_root, 'val')

# Create minimal directory structure
os.makedirs(traindir, exist_ok=True)
os.makedirs(valdir, exist_ok=True)

# Create a few dummy class directories and images
for class_idx in range(2):
    class_dir = os.path.join(traindir, f'class_{class_idx}')
    os.makedirs(class_dir, exist_ok=True)
    
    # Create a few dummy images
    for img_idx in range(5):
        img = Image.new('RGB', (256, 256), color=(73, 109, 137))
        img.save(os.path.join(class_dir, f'img_{img_idx}.jpg'))

ARGS = {
    'dataset_name': 'imagenet',
    'batch_size': 2,
    'n_worker': 0,
    'val_size': 4,
    'train_size': None,
    'random_seed': 1,
    'data_root': data_root,
    'for_inception': False,
    'shuffle': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch, torchvision — 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 float — 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
make_divisible Ran
ran, honoured the contract we drafted
mit-han-lab/haq-release/models/mobilenetv2.py
pointer only (licence: MIT); the stored copy get_code serves, from chris-boson/fashion_mnist, is pointer only (licence: NONE) · get_code("63ec083689f80f47")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from chris-boson/fashion_mnist, 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.

quantize Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("7d546c0310d1c591")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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.

quantize_grad Ran
ran, honoured the contract we drafted
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("f5f448783f152da8")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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.

sample_batch_indexes Ran
ran, honoured the contract we drafted
mit-han-lab/haq-release/lib/rl/memory.py
code served (permissive licence) · get_code("c83e49880d0e8a07")
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 list, 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 warnings
import random
import numpy as np

def sample_batch_indexes(low, high, size):
    if high - low >= size:
        try:
            r = xrange(low, high)
        except NameError:
            r = range(low, high)
        batch_idxs = random.sample(r, size)
    else:
        warnings.warn('Not enough entries to sample without replacement. Consider increasing your warm-up phase to avoid oversampling!')
        batch_idxs = np.random.random_integers(low, high - 1, size=size)
    assert len(batch_idxs) == size
    return batch_idxs

ARGS = {
    "low": 0,
    "high": 100,
    "size": 10
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family int — a list 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
zeroed_observation Ran
ran, honoured the contract we drafted
mit-han-lab/haq-release/lib/rl/memory.py
pointer only (licence: MIT); the stored copy get_code serves, from fshamshirdar/pytorch-rdpg, is pointer only (licence: NONE) · get_code("b5cc72675b688b00")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from fshamshirdar/pytorch-rdpg, 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.

QConv2d Not yet run
unverified (harvested, no recorded run)
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("756d13995db71fff")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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. No invocation recorded.

ResNetRecurrentGateSP Not yet run
unverified (harvested, no recorded run)
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("9839bc64a4774f55")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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. No invocation recorded.

UniformQuantize Not yet run
unverified (harvested, no recorded run)
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("1a2456136d59fff3")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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. No invocation recorded.

UniformQuantizeGrad Not yet run
unverified (harvested, no recorded run)
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("753ba41bd9a74e32")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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. No invocation recorded.

conv3x3 Not yet run
unverified (harvested, no recorded run)
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("e50cc160a435968b")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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. No invocation recorded.

get_dataset Not yet run
unverified (harvested, no recorded run)
mit-han-lab/haq-release/lib/utils/data_utils.py
code served (permissive licence) · get_code("850e0644d35c1376")
How it ran

no invocation recorded.

repackage_hidden Not yet run
unverified (harvested, no recorded run)
Torment123/DFS/models.py
pointer only (licence: NONE) · get_code("b6f41456f1266f92")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Torment123/DFS 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. No invocation recorded.

train Not yet run
unverified (harvested, no recorded run)
mit-han-lab/haq-release/rl_quantize.py
code served (permissive licence) · get_code("d0be8747b3078aa4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 24 samples lifted from this paper's linked repositories, 24 carry a function identity and 0 do not. Of the 24 distinct functions with an identity, 12 are also present in other papers whose code we hold (12 with at least one paper dated earlier, 1 the same month, 6 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.

FunctionAlso present inEarliest-datedOthers named
make_divisible 12 other papers
7 dated earlier, 1 the same month, 4 later · 1 only as an assembly brick
1503.03832 (2015-03)
FaceNet: A Unified Embedding for Face Recognition and Clustering
1801.04381 (2018-01)
MobileNetV2: Inverted Residuals and Linear Bottlenecks
1811.08383 (2018-11)
TSM: Temporal Shift Module for Efficient Video Understanding
and 9 more not named here
repackage_hidden
brick of this paper's assembly: helper
6 other papers
4 dated earlier, 2 later
1709.07432 (2017-09)
Dynamic Evaluation of Neural Sequence Models
1710.02224 (2017-10)
Dilated Recurrent Neural Networks
1805.04623 (2018-05)
Sharp Nearby, Fuzzy Far Away: How Neural Language Models Use Context
and 3 more not named here
conv_1x1_bn 4 other papers
3 dated earlier, 1 later
1503.03832 (2015-03)
FaceNet: A Unified Embedding for Face Recognition and Clustering
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
1905.02244 (2019-05)
Searching for MobileNetV3
and 1 more not named here
conv_bn 4 other papers
2 dated earlier, 2 later
1503.03832 (2015-03)
FaceNet: A Unified Embedding for Face Recognition and Clustering
1905.02244 (2019-05)
Searching for MobileNetV3
2009.09960 (2020-09)
Towards Fast, Accurate and Stable 3D Dense Face Alignment
and 1 more not named here
sample_batch_indexes 4 other papers
3 dated earlier, 1 later
1509.02971 (2015-09)
Continuous control with deep reinforcement learning
1511.05952 (2015-11)
Prioritized Experience Replay
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
and 1 more not named here
zeroed_observation 4 other papers
3 dated earlier, 1 later · 1 only as an assembly brick
1509.02971 (2015-09)
Continuous control with deep reinforcement learning
1511.05952 (2015-11)
Prioritized Experience Replay
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
and 1 more not named here
conv_1x1_bn 1 other paper
1 dated earlier
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
conv_bn 1 other paper
1 dated earlier
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
conv_dw 1 other paper
1 dated earlier
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
get_dataset 1 other paper
1 dated earlier
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
get_split_train_dataset 1 other paper
1 dated earlier
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
train 1 other paper
1 dated earlier
1812.00332 (2018-12)
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware

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

While increasingly deep networks are still in general desired for achieving state-of-the-art performance, for many specific inputs a simpler network might already suffice. Existing works exploited this observation by learning to skip convolutional layers in an input-dependent manner. However, we argue their binary decision scheme, i.e., either fully executing or completely bypassing one layer for a specific input, can be enhanced by introducing finer-grained, "softer" decisions. We therefore propose a Dynamic Fractional Skipping (DFS) framework. The core idea of DFS is to hypothesize layer-wise quantization (to different bitwidths) as intermediate "soft" choices to be made between fully utilizing and skipping a layer. For each input, DFS dynamically assigns a bitwidth to both weights and activations of each layer, where fully executing and skipping could be viewed as two "extremes" (i.e., full bitwidth and zero bitwidth). In this way, DFS can "fractionally" exploit a layer's expressive power during input-adaptive inference, enabling finer-grained accuracy-computational cost trade-offs. It presents a unified view to link input-adaptive layer skipping and input-adaptive hybrid quantization. Extensive experimental results demonstrate the superior tradeoff between computational cost and model expressive power (accuracy) achieved by DFS. More visualizations also indicate a smooth and consistent transition in the DFS behaviors, especially the learned choices between layer skipping and different quantizations when the total computational budgets vary, validating our hypothesis that layer quantization could be viewed as intermediate variants of layer skipping. Our source code and supplementary material are available at https://github.com/Torment123/DFS.

For agents

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

get_harvested_code_for_paper("2001.00705")
get_code_for_paper("2001.00705")
have("2001.00705")

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

curl https://syntology.ai/api/ran/2001.00705.json

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

[![Syntology run record](https://syntology.ai/api/ran/2001.00705.svg)](https://syntology.ai/paper/2001.00705)

Connect an agent — have() is free.