SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2603.00599 · 2026

Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger

Ming Zhang, Hao Peng, Philip Yu, Li Sun, Zhenhao Huang, Sen Su, Zhongtian Sun, Wenxin Jin

arXiv · PDF · Open in the Atlas

Code that ran

12 of 12 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; 2 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 10 failures of our instrument; 0 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/2603.00599.json · badge: /api/ran/2603.00599.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
Mingzhang21/HealHGNN canonical (classifier label)
classifier basis: census
12 ran · 0 no recorded run
FunctionStatusWhere it lives
ActivateModule Ran
ran; our contract draft was wrong, not the code
Mingzhang21/HealHGNN/HealHGNN/utils/train_utils.py
code served (permissive licence) · get_code("e56d64255e4ee03e")
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 ReLU. 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)

class TeLU(nn.Module):
    def forward(self, x):
        return x * torch.tanh(torch.exp(x))

def ActivateModule(act_str: str):
    if act_str == 'relu':
        return nn.ReLU()
    elif act_str == 'leaky_relu':
        return nn.LeakyReLU(negative_slope=0.2, inplace=True)
    elif act_str == 'tanh':
        return nn.Tanh()
    elif act_str == 'elu':
        return nn.ELU()
    elif act_str == 'gelu':
        return nn.GELU()
    elif act_str == 'softplus':
        return nn.Softplus()
    elif act_str == 'sigmoid':
        return nn.Sigmoid()
    elif act_str is None:
        return nn.Identity()
    elif act_str == 'telu':
        return TeLU()
    else:
        raise NotImplementedError

ARGS = {
    'act_str': 'relu'
}
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 ReLU — a ReLU output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ReLU 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
NormModule Ran
ran; our contract draft was wrong, not the code
Mingzhang21/HealHGNN/HealHGNN/utils/train_utils.py
code served (permissive licence) · get_code("ae2eb46615e36cd3")
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 LayerNorm. 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 = {
    'norm_str': 'ln',
    'dim': 4
}
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 LayerNorm — a LayerNorm output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a LayerNorm 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
cal_AUC_AP Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Mingzhang21/HealHGNN/HealHGNN/utils/eval_utils.py
code served (permissive licence) · get_code("bfb52d89fac57a1e")
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 numpy as np

torch_manual_seed_called = False
try:
    import torch
    torch.manual_seed(0)
    torch_manual_seed_called = True
except ImportError:
    pass

# Create binary classification scores and true labels
np.random.seed(0)

# scores: predicted probabilities/scores for positive class
scores = np.array([0.1, 0.4, 0.35, 0.8, 0.9, 0.2, 0.7, 0.6])

# trues: binary ground truth labels
trues = np.array([0, 0, 1, 1, 1, 0, 1, 0])

ARGS = {
    "scores": scores,
    "trues": trues,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, 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 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
cal_F1 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Mingzhang21/HealHGNN/HealHGNN/utils/eval_utils.py
code served (permissive licence) · get_code("8d30d89985b51939")
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 torch
import numpy as np
from sklearn.metrics import f1_score

torch.manual_seed(0)

# cal_F1 expects:
# - preds: predicted labels (array-like, 1D)
# - trues: true labels (array-like, 1D)

# Create small binary classification predictions and true labels
preds = np.array([0, 1, 1, 0, 1, 0])
trues = np.array([0, 1, 0, 0, 1, 1])

ARGS = {
    'preds': preds,
    'trues': trues
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, 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 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
cal_accuracy Ran
ran, honoured the contract we drafted
Mingzhang21/HealHGNN/HealHGNN/utils/eval_utils.py
code served (permissive licence) · get_code("9dc4838146d93e39")
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 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
import numpy as np

torch.manual_seed(0)

# Construct arguments for cal_accuracy(preds, trues)
# The function compares preds == trues element-wise, sums correct matches, and divides by length
# Both preds and trues should be comparable (same shape, comparable types)

# Create small tensors with batch size 2
preds = torch.tensor([0, 1])
trues = torch.tensor([0, 1])

ARGS = {
    "preds": preds,
    "trues": trues
}
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 Tensor, 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
count_parameters Ran
ran, honoured the contract we drafted
Mingzhang21/HealHGNN/HealHGNN/utils/utils.py
code served (permissive licence) · get_code("f6b944f50d3f15ae")
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 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.

import torch
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.linear2(x)
        return x

model = SimpleModel()

ARGS = {
    "model": model
}
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 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
create_logger Ran
ran; our contract draft was wrong, not the code
Mingzhang21/HealHGNN/HealHGNN/logger.py
code served (permissive licence) · get_code("f5c6682786e66416")
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 RootLogger. 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 logging
import tempfile
import os

# ARGS dict for create_logger function
ARGS = {
    'filepath': os.path.join(tempfile.gettempdir(), 'test_logger.log'),
    'colored': False,
    'debug': False
}
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 RootLogger — a RootLogger output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a RootLogger 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
get_split Ran
ran; our contract draft was wrong, not the code
Mingzhang21/HealHGNN/HealHGNN/utils/data_utils.py
code served (permissive licence) · get_code("373a389551605855")
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 RandomNodeSplit. 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
from torch_geometric.transforms import RandomNodeSplit, LargestConnectedComponents, Compose

torch.manual_seed(0)

ARGS = {
    "num_splits": 3,
    "num_val": 0.2,
    "num_test": 0.2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torch_geometric — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type RandomNodeSplit — a RandomNodeSplit output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a RandomNodeSplit 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
list2str Ran
ran; our contract draft was wrong, not the code
Mingzhang21/HealHGNN/HealHGNN/utils/config.py
code served (permissive licence) · get_code("9290ac465e02bd05")
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 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.

import torch

torch.manual_seed(0)

# list2str takes a list and converts it to a string
# It joins elements with '_' separator, with the last element having no trailing '_'
# The function accesses l[:-1] (all but last), l[-1] (last element)
# So l must be a list with at least 1 element

l = [1, 2, 3, 4]

ARGS = {
    'l': l
}
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 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
load_config Ran
ran; our contract draft was wrong, not the code
Mingzhang21/HealHGNN/HealHGNN/utils/config.py
code served (permissive licence) · get_code("5c3415d2970d7a4c")
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 DotDict. 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 tempfile
import os

class DotDict(dict):
    __getattr__ = dict.__getitem__
    __setattr__ = dict.__setitem__
    __delattr__ = dict.__delitem__

    def __init__(self, dct):
        super().__init__()
        for key, value in dct.items():
            if hasattr(value, 'keys'):
                value = DotDict(value)
            self[key] = value

def load_config(config_dict, json_file: str):
    with open(json_file, 'rt') as f:
        config_dict.update(json.load(f))
    configs = DotDict(config_dict)
    f.close()
    return configs

# Create a temporary JSON file with sample configuration
temp_dir = tempfile.mkdtemp()
json_file_path = os.path.join(temp_dir, "config.json")

sample_config = {
    "model_name": "test_model",
    "learning_rate": 0.001,
    "batch_size": 32,
    "nested_config": {
        "hidden_size": 128,
        "num_layers": 2
    }
}

with open(json_file_path, 'w') as f:
    json.dump(sample_config, f)

# Create the config_dict argument
config_dict = {
    "initial_param": "initial_value",
    "epochs": 10
}

ARGS = {
    "config_dict": config_dict,
    "json_file": json_file_path
}
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 DotDict — a DotDict 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
rand_train_test_idx Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Mingzhang21/HealHGNN/HealHGNN/utils/utils.py
code served (permissive licence) · get_code("5e264c272c383c0c")
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 dict. 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

torch.manual_seed(0)

# Create a label tensor with multiple classes for testing both balance=True and balance=False
# Using 10 samples with 3 classes
label = torch.tensor([0, 0, 0, 1, 1, 1, 2, 2, 2, 2])

# Define proportions
train_prop = 0.6
valid_prop = 0.2

ARGS = {
    'label': label,
    'train_prop': train_prop,
    'valid_prop': valid_prop,
    'balance': False
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
save_data_to_pickle Ran
ran; our contract draft was wrong, not the code
Mingzhang21/HealHGNN/HealHGNN/utils/data_utils.py
code served (permissive licence) · get_code("95e21e1aa38d7b73")
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 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.

import os
import os.path as osp
import pickle
import tempfile

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

# Prepare test data - a simple dictionary with some data
test_data = {
    'nodes': [1, 2, 3, 4, 5],
    'edges': [(0, 1), (1, 2), (2, 3)],
    'features': [[0.1, 0.2], [0.3, 0.4], [0.5, 0.6]],
    'labels': [0, 1, 0, 1, 1]
}

ARGS = {
    'data': test_data,
    'p2root': temp_dir,
    'file_name': 'test_hypergraph_data'
}
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 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

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 9 are also present in other papers whose code we hold (9 with at least one paper dated earlier, 1 the same month, 1 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.

FunctionAlso present inEarliest-datedOthers named
count_parameters 180 other papers
171 dated earlier, 1 the same month, 8 later · 8 only as an assembly brick
1705.08415 (2017-05)
Supervised Community Detection with Line Graph Neural Networks
1707.01926 (2017-07)
Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting
1808.09367 (2018-08)
Deriving Machine Attention from Human Rationales
and 177 more not named here
cal_AUC_AP 3 other papers
3 dated earlier
2410.17941 (2024-10)
Spiking Graph Neural Network on Riemannian Manifolds
2412.11085 (2024-12)
GraphMoRE: Mitigating Topological Heterogeneity via Mixture of Riemannian Experts
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
create_logger 3 other papers
3 dated earlier
2410.17941 (2024-10)
Spiking Graph Neural Network on Riemannian Manifolds
2412.11085 (2024-12)
GraphMoRE: Mitigating Topological Heterogeneity via Mixture of Riemannian Experts
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
list2str 2 other papers
2 dated earlier
2410.17941 (2024-10)
Spiking Graph Neural Network on Riemannian Manifolds
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
load_config 2 other papers
2 dated earlier
2410.17941 (2024-10)
Spiking Graph Neural Network on Riemannian Manifolds
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
NormModule 1 other paper
1 dated earlier
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
cal_F1 1 other paper
1 dated earlier
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
cal_accuracy 1 other paper
1 dated earlier
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
get_split 1 other paper
1 dated earlier
2510.17457 (2025-10)
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models

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

Hypergraphs are the natural description of higher-order interactions among objects, widely applied in social network analysis, cross-modal retrieval, etc. Hypergraph Neural Networks (HGNNs) have become the dominant solution for learning on hypergraphs. Traditional HGNNs are extended from message passing graph neural networks, following the homophily assumption, and thus struggle with the prevalent heterophilic hypergraphs that call for longrange dependence modeling. Existing solutions enlarge the message flow through the hypergraph bottleneck, mitigating the oversquashing issue and capturing long-range dependence. However, they often accelerate the loss of representation distinguishability in the repeated aggregations, leading to oversmoothing. This dilemma motivates an interesting question: Can we develop a unified mechanism that is agnostic to both homophilic and heterophilic hypergraphs? In this paper, we achieve the best of both worlds through the lens of Riemannian geometry, which provides the potential to adjust the message passing behavior in different regions. The key insight lies in the connection between oversquashing and hypergraph bottleneck within the framework of Riemannian manifold heat flow. Building on this, we propose the novel idea of locally adapting the bottlenecks of different subhypergraphs. The core innovation of the proposed mechanism is the design of an adaptive local (heat) exchanger. Specifically, it captures the rich long-range dependencies via the Robin condition, and preserves the representation distinguishability via source terms, * Corresponding Author.

For agents

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

get_harvested_code_for_paper("2603.00599")
get_code_for_paper("2603.00599")
have("2603.00599")

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

curl https://syntology.ai/api/ran/2603.00599.json

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

[![Syntology run record](https://syntology.ai/api/ran/2603.00599.svg)](https://syntology.ai/paper/2603.00599)

Connect an agent — have() is free.