SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.13072 · NeurIPS · 2023

Interpretable Mesomorphic Neural Networks For Tabular Data

Josif Grabocka, Sebastian Pineda, Arlind Kadra

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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 0 have no contract verdict; 3 failures of our instrument; 0 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/2305.13072.json · badge: /api/ran/2305.13072.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
arlindkadra/imn canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
calculate_method_time Ran
ran; our contract draft was wrong, not the code
arlindkadra/imn/plots/comparison.py
code served (permissive licence) · get_code("87aa64b96c8ddbf5")
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 DataFrame. 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 json
import os
import tempfile
import shutil

# Create a temporary directory structure for testing
temp_dir = tempfile.mkdtemp()

# Create the directory structure: output_dir/method_name/dataset_id/seed/output_info.json
method_name = "test_method"
dataset_id = "dataset_001"
seed = "seed_0"

method_output_dir = os.path.join(temp_dir, method_name)
dataset_dir = os.path.join(method_output_dir, dataset_id)
seed_dir = os.path.join(dataset_dir, seed)

os.makedirs(seed_dir, exist_ok=True)

# Create output_info.json files with timing data
output_info = {
    "train_time": 10.5,
    "inference_time": 2.3
}

with open(os.path.join(seed_dir, 'output_info.json'), 'w') as f:
    json.dump(output_info, f)

# Create another seed for the same dataset to test median calculation
seed_2 = "seed_1"
seed_dir_2 = os.path.join(dataset_dir, seed_2)
os.makedirs(seed_dir_2, exist_ok=True)

output_info_2 = {
    "train_time": 11.5,
    "inference_time": 2.5
}

with open(os.path.join(seed_dir_2, 'output_info.json'), 'w') as f:
    json.dump(output_info_2, f)

# Create another dataset
dataset_id_2 = "dataset_002"
dataset_dir_2 = os.path.join(method_output_dir, dataset_id_2)
seed_dir_3 = os.path.join(dataset_dir_2, "seed_0")
os.makedirs(seed_dir_3, exist_ok=True)

output_info_3 = {
    "train_time": 15.0,
    "inference_time": 3.0
}

with open(os.path.join(seed_dir_3, 'output_info.json'), 'w') as f:
    json.dump(output_info_3, f)

ARGS = {
    "output_dir": temp_dir,
    "method_name": method_name
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
calculate_nr_trials Ran
ran; our contract draft was wrong, not the code
arlindkadra/imn/plots/comparison.py
code served (permissive licence) · get_code("5d1f2edc21eb7903")
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 DataFrame. 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 os
import tempfile
import pandas as pd

# Create a temporary directory structure that the function expects
temp_dir = tempfile.mkdtemp()

# Create the directory structure: output_dir/method_name/dataset_id/0/trials.csv
method_name = "test_method"
dataset_id = "dataset_001"

method_output_dir = os.path.join(temp_dir, method_name)
dataset_dir = os.path.join(method_output_dir, dataset_id)
seed_dir = os.path.join(dataset_dir, "0")

os.makedirs(seed_dir, exist_ok=True)

# Create a sample trials.csv file
trials_data = pd.DataFrame({
    'trial_id': [1, 2, 3],
    'loss': [0.5, 0.4, 0.3],
    'accuracy': [0.7, 0.8, 0.9]
})
trials_csv_path = os.path.join(seed_dir, 'trials.csv')
trials_data.to_csv(trials_csv_path, index=False)

# Create another dataset to test multiple datasets
dataset_id_2 = "dataset_002"
dataset_dir_2 = os.path.join(method_output_dir, dataset_id_2)
seed_dir_2 = os.path.join(dataset_dir_2, "0")
os.makedirs(seed_dir_2, exist_ok=True)

trials_data_2 = pd.DataFrame({
    'trial_id': [1, 2],
    'loss': [0.6, 0.5],
    'accuracy': [0.65, 0.75]
})
trials_csv_path_2 = os.path.join(seed_dir_2, 'trials.csv')
trials_data_2.to_csv(trials_csv_path_2, index=False)

ARGS = {
    'output_dir': temp_dir,
    'method_name': method_name
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
prepare_method_results Ran
ran; our contract draft was wrong, not the code
arlindkadra/imn/plots/comparison.py
code served (permissive licence) · get_code("3371ef8337480dbe")
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 DataFrame. 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 json
import os
import tempfile
import shutil

# Create a temporary directory structure for testing
temp_dir = tempfile.mkdtemp()

# Create the directory structure: output_dir/method_name/dataset_id/seed/
method_name = "test_method"
dataset_id = "dataset_1"
seed = "seed_0"

method_output_dir = os.path.join(temp_dir, method_name)
dataset_dir = os.path.join(method_output_dir, dataset_id)
seed_dir = os.path.join(dataset_dir, seed)

os.makedirs(seed_dir, exist_ok=True)

# Create output_info.json files with test and train auroc values
output_info = {
    "test_auroc": 0.85,
    "train_auroc": 0.90
}

with open(os.path.join(seed_dir, 'output_info.json'), 'w') as f:
    json.dump(output_info, f)

# Create another seed for the same dataset
seed_2 = "seed_1"
seed_dir_2 = os.path.join(dataset_dir, seed_2)
os.makedirs(seed_dir_2, exist_ok=True)

output_info_2 = {
    "test_auroc": 0.87,
    "train_auroc": 0.92
}

with open(os.path.join(seed_dir_2, 'output_info.json'), 'w') as f:
    json.dump(output_info_2, f)

# Create another dataset
dataset_id_2 = "dataset_2"
dataset_dir_2 = os.path.join(method_output_dir, dataset_id_2)
seed_dir_3 = os.path.join(dataset_dir_2, "seed_0")
os.makedirs(seed_dir_3, exist_ok=True)

output_info_3 = {
    "test_auroc": 0.80,
    "train_auroc": 0.88
}

with open(os.path.join(seed_dir_3, 'output_info.json'), 'w') as f:
    json.dump(output_info_3, f)

ARGS = {
    "output_dir": temp_dir,
    "method_name": method_name
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 3 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.

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

Even though neural networks have been long deployed in applications involving tabular data, still existing neural architectures are not explainable by design. In this work, we propose a new class of interpretable neural networks for tabular data that are both deep and linear at the same time (i.e. mesomorphic). We optimize deep hypernetworks to generate explainable linear models on a per-instance basis. As a result, our models retain the accuracy of black-box deep networks while offering free lunch explainability for tabular data by design. Through extensive experiments, we demonstrate that our explainable deep networks have comparable performance to state-of-the-art classifiers on tabular data and outperform current existing methods that are explainable by design.

For agents

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

get_harvested_code_for_paper("2305.13072")
get_code_for_paper("2305.13072")
have("2305.13072")

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

curl https://syntology.ai/api/ran/2305.13072.json

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

[![Syntology run record](https://syntology.ai/api/ran/2305.13072.svg)](https://syntology.ai/paper/2305.13072)

Connect an agent — have() is free.