SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.03348 · ICLR · 2024

Respect the model: Fine-grained and Robust Explanation with Sharing Ratio Decomposition

Nojun Kwak, Sangyu Han, Yearim Kim

arXiv · PDF · Open in the Atlas

Code that ran

3 of 11 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; 1 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 failures of our instrument; 8 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/2402.03348.json · badge: /api/ran/2402.03348.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
Sangyu-Han/SharingRatioDecomposition community (classifier label)
classifier basis not recorded on the link
3 ran · 8 no recorded run
FunctionStatusWhere it lives
enlarge_image Ran
ran on a synthesized input
Sangyu-Han/SharingRatioDecomposition/render.py
code served (permissive licence) · get_code("65c3ef82157a96e7")
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 ndarray, ndim 2, dtype family float, shape (6, 9) 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 numpy as np

# Create a 2D test image (grayscale)
img_2d = np.array([[1, 2, 3],
                   [4, 5, 6]], dtype=np.float32)

# Create a 3D test image (RGB)
img_3d = np.array([[[10, 20, 30],
                    [40, 50, 60]],
                   [[70, 80, 90],
                    [100, 110, 120]]], dtype=np.float32)

ARGS = {
    "img": img_2d,
    "scaling": 3
}
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 ndarray, ndim 2, dtype family float, shape (6, 9) — shape (6, 9) 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
repaint_corner_pixels Ran
ran on a synthesized input
Sangyu-Han/SharingRatioDecomposition/render.py
code served (permissive licence) · get_code("641ea90c3a790631")
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 ndarray, ndim 3, dtype family float, shape (10, 8, 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 numpy as np

# Create a small RGB image with shape [H x W x 3]
# Using 10x8 dimensions for height and width to have distinct sizes
H, W = 10, 8
rgbimg = np.random.randint(0, 256, size=(H, W, 3), dtype=np.uint8).astype(np.float32)

# Set some specific values for corner pixels to make the function behavior clear
rgbimg[0, 3, :] = [100, 150, 200]      # rgbimg[0, scaling, :]
rgbimg[3, 0, :] = [50, 100, 150]       # rgbimg[scaling, 0, :]
rgbimg[3, 3, :] = [75, 125, 175]       # rgbimg[scaling, scaling, :]

rgbimg[-1, -4, :] = [200, 100, 50]     # rgbimg[-1, -1-scaling, :]
rgbimg[-4, -1, :] = [150, 100, 50]     # rgbimg[-1-scaling, -1, :]
rgbimg[-4, -4, :] = [175, 125, 75]     # rgbimg[-1-scaling, -1-scaling, :]

ARGS = {
    "rgbimg": rgbimg,
    "scaling": 3
}
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 ndarray, ndim 3, dtype family float, shape (10, 8, 3) — shape (10, 8, 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
safe_divide Ran
ran, honoured the contract we drafted
Sangyu-Han/SharingRatioDecomposition/modules/layers.py
pointer only (licence: MIT); the stored copy get_code serves, from wjNam/Relative_Attributing_Propagation, is pointer only (licence: NONE) · get_code("251e368f5f9b12f5")
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 wjNam/Relative_Attributing_Propagation, 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.

cal_simmap_cbr Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/util.py
code served (permissive licence) · get_code("33e7798cbdceb96f")
How it ran

no invocation recorded.

cal_simmap_input_cr Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/util.py
code served (permissive licence) · get_code("749b3c61b3663b4d")
How it ran

no invocation recorded.

cal_simmap_x Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/util.py
code served (permissive licence) · get_code("8aef1815e26420a0")
How it ran

no invocation recorded.

conv1x1 Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/modules/resnet.py
code served (permissive licence) · get_code("45b3400cd186760f")
How it ran

no invocation recorded.

conv3x3 Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/modules/resnet.py
code served (permissive licence) · get_code("e14e792923db61dd")
How it ran

no invocation recorded.

make_layers Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/modules/vgg.py
code served (permissive licence) · get_code("506e0e8c845fe9c2")
How it ran

no invocation recorded.

resnet18 Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/modules/resnet.py
code served (permissive licence) · get_code("8a77bc582a4791c2")
How it ran

no invocation recorded.

vec2im Not yet run
unverified (harvested, no recorded run)
Sangyu-Han/SharingRatioDecomposition/render.py
code served (permissive licence) · get_code("db2fd24be9a8e381")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 1 the same month, 2 later) and 9 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
resnet18 89 other papers
73 dated earlier, 2 the same month, 12 later, 2 same year, order not determinable
1705.09759 (2017-05)
CASENet: Deep Category-Aware Semantic Edge Detection
1801.07698 (2018-01)
ArcFace: Additive Angular Margin Loss for Deep Face Recognition
1803.00839 (2018-03)
Pose-Robust Face Recognition via Deep Residual Equivariant Mapping
and 86 more not named here
safe_divide 3 other papers
2 dated earlier, 1 later
1904.00605 (2019-04)
Relative Attributing Propagation: Interpreting the Comparative Contributions of Individual Units in Deep Neural Networks
2310.04519 (2023-10)
SPADE: Sparsity-Guided Debugging for Deep Neural Networks
2403.09274 (2024-03)
EventRPG: Event Data Augmentation with Relevance Propagation Guidance

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

The truthfulness of existing explanation methods in authentically elucidating the underlying model's decision-making process has been questioned. Existing methods have deviated from faithfully representing the model, thus susceptible to adversarial attacks. To address this, we propose a novel eXplainable AI (XAI) method called SRD (Sharing Ratio Decomposition), which sincerely reflects the model's inference process, resulting in significantly enhanced robustness in our explanations. Different from the conventional emphasis on the neuronal level, we adopt a vector perspective to consider the intricate nonlinear interactions between filters. We also introduce an interesting observation termed Activation-Pattern-Only Prediction (APOP), letting us emphasize the importance of inactive neurons and redefine relevance encapsulating all relevant information including both active and inactive neurons. Our method, SRD, allows for the recursive decomposition of a Pointwise Feature Vector (PFV), providing a high-resolution Effective Receptive Field (ERF) at any layer. * Equal contribution. † Corresponding author. 1 A neuron outputs a scalar, an element in a tensor, by combining the information in its receptive field.

For agents

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

get_harvested_code_for_paper("2402.03348")
get_code_for_paper("2402.03348")
have("2402.03348")

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

curl https://syntology.ai/api/ran/2402.03348.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.03348.svg)](https://syntology.ai/paper/2402.03348)

Connect an agent — have() is free.