SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2409.02335 · ICLR · 2025

What Do You See in Common? Learning Hierarchical Prototypes over Tree-of-Life to Discover Evolutionary Traits

Wei-Lun Chao, Arka Daw, M Maruf, Anuj Karpatne, Tanya Berger-Wolf, Kaiya Provost, Harish Manogaran, Kazi Mehrab, Caleb Charpentier, Josef Uyeda, Wasila Dahdul, Matthew Thompson, and 3 more

arXiv · PDF · Open in the Atlas

Code that ran

8 of 15 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; 3 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 4 failures of our instrument; 7 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/2409.02335.json · badge: /api/ran/2409.02335.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
imageomics/hcompnet canonical (classifier label)
classifier basis: census
4 ran · 6 no recorded run
Imageomics/HComPNet canonical (classifier label)
classifier basis: classifier_2025
4 ran · 1 no recorded run
FunctionStatusWhere it lives
alter_name_for_cub_phylogeny Ran
ran; our contract draft was wrong, not the code
Imageomics/HComPNet/preprocess_data/prepare_cub190.py
code served (permissive licence) · get_code("589a58686081bfab")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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.

ARGS = {
    'name': '123_456'
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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
apply_segmentation_mask_with_imagenet_mean Ran
ran, honoured the contract we drafted
Imageomics/HComPNet/preprocess_data/prepare_cub190.py
code served (permissive licence) · get_code("133c7b2a2ef12d5c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned Image. 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 numpy as np
from PIL import Image
import io
import tempfile
import os

# Create a small test image (RGBA)
image_array = np.random.randint(0, 255, (8, 6, 3), dtype=np.uint8)
image = Image.fromarray(image_array, mode='RGB')

# Create a small segmentation mask (grayscale)
mask_array = np.random.randint(0, 255, (8, 6), dtype=np.uint8)
mask = Image.fromarray(mask_array, mode='L')

# Save to temporary files
temp_dir = tempfile.gettempdir()
image_path = os.path.join(temp_dir, 'test_image.png')
mask_path = os.path.join(temp_dir, 'test_mask.png')

image.save(image_path)
mask.save(mask_path)

ARGS = {
    'image_path': image_path,
    'segmentation_mask_path': mask_path
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
PIL, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
exclude_bias_and_batchnorm Ran
ran; our contract draft was wrong, not the code
Imageomics/HComPNet/util/args.py
code served (permissive licence) · get_code("752d0617b8364e73")
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. 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 simple model with bias, batchnorm, and regular parameters
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.bn1 = nn.BatchNorm1d(3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.bn1(x)
        x = self.linear2(x)
        return x

model = SimpleModel()

# Create named_params dict from model's named_parameters()
named_params = dict(model.named_parameters())

ARGS = {
    'named_params': named_params
}
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 — a tuple output has no array shape of its own; none is 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_patch_size Ran
ran, honoured the contract we drafted
imageomics/hcompnet/util/func.py
code served (permissive licence) · get_code("fc27da95703ce673")
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 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
import numpy as np

class Args:
    def __init__(self):
        self.image_size = 224
        self.wshape = 7

ARGS = {"args": Args()}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, 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
is_bias_or_batchnorm Ran
ran, violated the contract we drafted
Imageomics/HComPNet/util/args.py
code served (permissive licence) · get_code("f0fa33b411bf906e")
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 bool, 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 torch

torch.manual_seed(0)

ARGS = {
    'param_name': 'layer.bias'
}
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 bool, 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
make_squared Ran
ran; our contract draft was wrong, not the code
imageomics/hcompnet/preprocess_data/prepare_cub190.py
code served (permissive licence) · get_code("3f4ca34ecc5188e9")
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 Image. 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 numpy as np
from PIL import Image
from torchvision import transforms

class Args:
    def __init__(self, image_size):
        self.image_size = image_size

# Create a small test image (PIL Image)
# Make it rectangular (not square) to test the resizing logic
test_img = Image.new('RGB', (100, 60), color=(200, 150, 100))

ARGS = {
    'args': Args(image_size=128),
    'img': test_img,
    'padding': 'white'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
replace_convlayers_convnext Ran
ran; our contract draft was wrong, not the code
imageomics/hcompnet/features/convnext_features.py
code served (permissive licence) · get_code("a758c95ddd9cf6e6")
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 SimpleModel. 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)

# Build a model with Conv2d layers at various depths
# The function recursively traverses named_children and checks for Conv2d instances
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        # Conv2d with stride=2 and in_channels > threshold
        self.conv1 = nn.Conv2d(in_channels=16, out_channels=32, kernel_size=3, stride=2, padding=1)
        # Conv2d with stride=2 but in_channels <= threshold
        self.conv2 = nn.Conv2d(in_channels=8, out_channels=16, kernel_size=3, stride=2, padding=1)
        # Conv2d with stride=1 (should not be modified)
        self.conv3 = nn.Conv2d(in_channels=32, out_channels=64, kernel_size=3, stride=1, padding=1)
        # Nested module with Conv2d
        self.nested = nn.Sequential(
            nn.Conv2d(in_channels=20, out_channels=40, kernel_size=3, stride=2, padding=1)
        )

model = SimpleModel()
threshold = 10

ARGS = {
    "model": model,
    "threshold": threshold,
}
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 SimpleModel — a SimpleModel output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModel 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
split_value Ran
ran, honoured the contract we drafted
imageomics/hcompnet/util/node.py
code served (permissive licence) · get_code("8a56e2bc50b4bd55")
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 torch
import numpy as np

torch.manual_seed(0)

# Analyze the function:
# split_value(M, n) takes two parameters:
# - M: used in divmod(M, n), so should be an integer
# - n: used in divmod(M, n) and range(remainder), so should be a positive integer
# The function splits M into n parts as evenly as possible

ARGS = {
    "M": 17,
    "n": 5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — 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
convnext_tiny_13_features Not yet run
unverified (harvested, no recorded run)
imageomics/hcompnet/features/convnext_features.py
code served (permissive licence) · get_code("094b66474a792977")
How it ran

no invocation recorded.

convnext_tiny_26_features Not yet run
unverified (harvested, no recorded run)
imageomics/hcompnet/features/convnext_features.py
code served (permissive licence) · get_code("c3c86ab4d3337a4f")
How it ran

no invocation recorded.

create_filtered_dataloader Not yet run
unverified (harvested, no recorded run)
imageomics/hcompnet/util/data.py
code served (permissive licence) · get_code("cda3521af97352b1")
How it ran

no invocation recorded.

get_data Not yet run
unverified (harvested, no recorded run)
imageomics/hcompnet/util/data.py
code served (permissive licence) · get_code("f07e5e822b7d8619")
How it ran

no invocation recorded.

get_optimizer_nn Not yet run
unverified (harvested, no recorded run)
Imageomics/HComPNet/util/args.py
code served (permissive licence) · get_code("77b218509c30e14c")
How it ran

no invocation recorded.

init_weights_xavier Not yet run
unverified (harvested, no recorded run)
imageomics/hcompnet/util/func.py
code served (permissive licence) · get_code("4195aead1609eeb2")
How it ran

no invocation recorded.

unshuffle_dataloader Not yet run
unverified (harvested, no recorded run)
imageomics/hcompnet/util/data.py
code served (permissive licence) · get_code("4d3d935f60f0e4a1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 0 later) and 11 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
convnext_tiny_13_features 1 other paper
1 dated earlier
2307.10404 (2023-07)
Interpreting and Correcting Medical Image Classification with PIP-Net
convnext_tiny_26_features 1 other paper
1 dated earlier
2307.10404 (2023-07)
Interpreting and Correcting Medical Image Classification with PIP-Net
get_patch_size 1 other paper
1 dated earlier
2307.10404 (2023-07)
Interpreting and Correcting Medical Image Classification with PIP-Net
replace_convlayers_convnext 1 other paper
1 dated earlier
2307.10404 (2023-07)
Interpreting and Correcting Medical Image Classification with PIP-Net

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.

For agents

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

get_harvested_code_for_paper("2409.02335")
get_code_for_paper("2409.02335")
have("2409.02335")

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

curl https://syntology.ai/api/ran/2409.02335.json

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

[![Syntology run record](https://syntology.ai/api/ran/2409.02335.svg)](https://syntology.ai/paper/2409.02335)

Connect an agent — have() is free.