SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.06757 · 2020

Relational Message Passing for Knowledge Graph Completion

arXiv · PDF · Open in the Atlas

Code that ran

12 of 18 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 10 have no contract verdict; 1 failure of our instrument; 6 no recorded run), measured 2026-11-21; 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/2002.06757.json · badge: /api/ran/2002.06757.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
muhanzhang/IGMC canonical (classifier label)
classifier basis: census
8 ran · 3 no recorded run
hwwang55/PathCon canonical (classifier label)
classifier basis: census
4 ran · 3 no recorded run
FunctionStatusWhere it lives
construct_pyg_graph Ran
ran on a synthesized input
muhanzhang/IGMC/util_functions.py
code served (permissive licence) · get_code("e673939466b977db")
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 Data. 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 numpy as np
from torch_geometric.data import Data

torch.manual_seed(0)
np.random.seed(0)

# Define the one_hot function from the module
def one_hot(idx, length):
    idx = np.array(idx)
    x = np.zeros([len(idx), length])
    x[np.arange(len(idx)), idx] = 1.0
    return x

ARGS = {
    "u": [0, 1],
    "v": [1, 0],
    "r": [0, 1],
    "node_labels": [0, 1],
    "max_node_label": 2,
    "y": 1.0,
    "node_features": None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, torch_geometric — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Data — a Data 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
count_all_paths Ran
ran on a synthesized input
hwwang55/PathCon/src/utils.py
code served (permissive licence) · get_code("43b5e3fa582da3d7")
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.

from collections import defaultdict

# Build e2re: entity to relationship-entity pairs mapping
# e2re[entity] = [(relation_id, target_entity), ...]
e2re = {
    0: [(0, 1), (1, 2)],      # entity 0 connects to entities 1 and 2
    1: [(2, 3), (3, 0)],      # entity 1 connects to entities 3 and 0
    2: [(4, 3)],              # entity 2 connects to entity 3
    3: [(5, 0), (6, 1)],      # entity 3 connects to entities 0 and 1
}

# Build head2tails: list of (head_entity, set_of_tail_entities)
head2tails = [
    (0, {3}),      # from head 0, we want paths to tail 3
    (1, {0, 2}),   # from head 1, we want paths to tails 0 and 2
]

max_path_len = 3
pid = 1

inputs = (e2re, max_path_len, head2tails, pid)

ARGS = {
    "inputs": inputs
}
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 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
get_layer_id Ran
ran on a synthesized input
hwwang55/PathCon/src/aggregators.py
code served (permissive licence) · get_code("fc9ffc41618b3fd4")
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 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 numpy as np

torch.manual_seed(0)

# The function get_layer_id takes one parameter: aggregator_name (a string)
# It uses the aggregator_name as a key in the AGGREGATOR_ID dict
# The function can be called with or without the parameter (has default value '')

ARGS = {
    'aggregator_name': 'test_aggregator'
}
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 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
get_params_for_mp Ran
ran on a synthesized input
hwwang55/PathCon/src/utils.py
code served (permissive licence) · get_code("b5db0d52d633b48d")
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. 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 multiprocessing as mp

def get_params_for_mp(n_triples):
    n_cores = mp.cpu_count()
    pool = mp.Pool(n_cores)
    avg = n_triples // n_cores
    range_list = []
    start = 0
    for i in range(n_cores):
        num = avg + 1 if i < n_triples - avg * n_cores else avg
        range_list.append([start, start + num])
        start += num
    return (n_cores, pool, range_list)

ARGS = {
    "n_triples": 100
}
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 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
links2subgraphs Ran
ran on a synthesized input
muhanzhang/IGMC/util_functions.py
code served (permissive licence) · get_code("671a634f85ff02bf")
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. 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 numpy as np
import scipy.sparse as ssp
from torch_geometric.data import Data

torch.manual_seed(0)
np.random.seed(0)

# Create a small sparse adjacency matrix (CSR format)
# 5 nodes on each side
n_u = 5
n_v = 5
rows = [0, 1, 2, 3, 4, 0, 1, 2]
cols = [0, 1, 2, 3, 4, 2, 3, 4]
data = [1, 1, 1, 1, 1, 1, 1, 1]
Arow = ssp.csr_matrix((data, (rows, cols)), shape=(n_u, n_v))
Acol = ssp.csr_matrix((data, (cols, rows)), shape=(n_v, n_u))

# Create links (edge pairs to extract subgraphs around)
links = ([0, 1, 2], [1, 2, 3])

# Create labels for each link
labels = [0, 1, 0]

# Create u and v features (small dense arrays)
u_features = np.random.randn(n_u, 4).astype(np.float32)
v_features = np.random.randn(n_v, 4).astype(np.float32)

# Class values mapping
class_values = {0: 0.0, 1: 1.0}

ARGS = {
    'Arow': Arow,
    'Acol': Acol,
    'links': links,
    'labels': labels,
    'h': 1,
    'sample_ratio': 1.0,
    'max_nodes_per_hop': None,
    'u_features': u_features,
    'v_features': v_features,
    'class_values': class_values,
    'parallel': False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, torch, torch_geometric, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list 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
load_matlab_file Ran
ran on a synthesized input
muhanzhang/IGMC/preprocessing.py
code served (permissive licence) · get_code("76472b9c1b3e64e0")
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 csc_matrix. 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
import scipy.sparse as sp
import h5py
import tempfile
import os

# Create a temporary .mat file in HDF5 format for testing
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.mat")

# Create a test HDF5 file with sparse matrix data
with h5py.File(temp_file, 'w') as f:
    # Create a sparse matrix group
    group = f.create_group('sparse_matrix')
    
    # Create sparse matrix data (CSC format)
    data = np.array([1.0, 2.0, 3.0, 4.0], dtype=np.float32)
    ir = np.array([0, 2, 1, 3], dtype=np.int32)  # row indices
    jc = np.array([0, 2, 4], dtype=np.int32)     # column pointers
    
    group.create_dataset('data', data=data)
    group.create_dataset('ir', data=ir)
    group.create_dataset('jc', data=jc)

# Create a test HDF5 file with dense matrix data
temp_file_dense = os.path.join(temp_dir, "test_dense.mat")
with h5py.File(temp_file_dense, 'w') as f:
    dense_data = np.array([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]], dtype=np.float32)
    f.create_dataset('dense_matrix', data=dense_data)

ARGS = {
    'path_file': temp_file,
    'name_field': 'sparse_matrix'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
h5py, numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type csc_matrix — a csc_matrix output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a csc_matrix 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
map_data Ran
ran; our contract draft was wrong, not the code
muhanzhang/IGMC/data_utils.py
code served (permissive licence) · get_code("c5610bcfc7d8a106")
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. 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

# Create test data: an array of int32 with non-continuous indices
data = np.array([5, 10, 5, 3, 10, 3, 5], dtype=np.int32)

ARGS = {
    "data": data
}
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 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
normalize_features Ran
ran on a synthesized input
muhanzhang/IGMC/preprocessing.py
code served (permissive licence) · get_code("1e9e875e5b3efe21")
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 csr_matrix. 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
import scipy.sparse as sp

# Create a small sparse matrix for testing
# The function expects a sparse matrix with:
# - .sum(1) method (row sums)
# - .dot() method (matrix multiplication)
# - .nnz attribute (number of non-zero elements)

# Create a 4x3 sparse matrix with some non-zero entries
row = np.array([0, 1, 2, 3, 0, 2])
col = np.array([0, 1, 2, 0, 2, 1])
data = np.array([1.0, 2.0, 3.0, 1.0, 2.0, 1.0])
feat = sp.csr_matrix((data, (row, col)), shape=(4, 3))

ARGS = {
    'feat': feat
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type csr_matrix — a csr_matrix output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a csr_matrix 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
num_graphs Ran
ran, honoured the contract we drafted
muhanzhang/IGMC/train_eval.py
pointer only (licence: MIT); the stored copy get_code serves, from RingBDStack/VIB-GSL, is pointer only (licence: NONE) · get_code("9da87b0b1605a660")
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 RingBDStack/VIB-GSL, 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.

preprocess_user_item_features Ran
ran on a synthesized input
muhanzhang/IGMC/preprocessing.py
code served (permissive licence) · get_code("0bb0ea10173803fd")
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 scipy.sparse as sp
import numpy as np

# Create small sparse matrices for user and item features
# u_features: (num_users, user_feature_dim)
# v_features: (num_items, item_feature_dim)
num_users = 2
num_items = 3
user_feature_dim = 4
item_feature_dim = 5

# Create sparse matrices with some non-zero values
u_data = np.array([1.0, 2.0, 3.0, 4.0])
u_row = np.array([0, 0, 1, 1])
u_col = np.array([0, 1, 2, 3])
u_features = sp.csr_matrix((u_data, (u_row, u_col)), shape=(num_users, user_feature_dim), dtype=np.float32)

v_data = np.array([5.0, 6.0, 7.0, 8.0, 9.0])
v_row = np.array([0, 1, 1, 2, 2])
v_col = np.array([0, 1, 2, 3, 4])
v_features = sp.csr_matrix((v_data, (v_row, v_col)), shape=(num_items, item_feature_dim), dtype=np.float32)

ARGS = {
    'u_features': u_features,
    'v_features': v_features,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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
read_entities Ran
ran on a synthesized input
hwwang55/PathCon/src/data_loader.py
code served (permissive licence) · get_code("fd92338302edcbb5")
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 tempfile
import os

# Create a temporary file with entity data
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("0\tentity_one\n")
temp_file.write("1\tentity_two\n")
temp_file.write("2\tentity_three\n")
temp_file.close()

ARGS = {
    'file_name': temp_file.name
}
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 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
subgraph_extraction_labeling Ran
ran on a synthesized input
muhanzhang/IGMC/util_functions.py
code served (permissive licence) · get_code("4308078bf13d0be4")
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. 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
import scipy.sparse as ssp

# Create a small bipartite graph represented as sparse matrices
# Arow: adjacency from u-nodes perspective (u_nodes x v_nodes)
# Acol: adjacency from v-nodes perspective (v_nodes x u_nodes)

# Create Arow as a sparse matrix (5 u-nodes x 4 v-nodes)
row_indices = np.array([0, 0, 1, 2, 3, 4])
col_indices = np.array([0, 1, 1, 2, 2, 3])
data = np.array([1, 1, 1, 1, 1, 1])
Arow = ssp.csr_matrix((data, (row_indices, col_indices)), shape=(5, 4))

# Create Acol as the transpose (4 v-nodes x 5 u-nodes)
Acol = Arow.transpose().tocsr()

# Create feature matrices
u_features = np.random.randn(5, 3)  # 5 u-nodes, 3 features
v_features = np.random.randn(4, 2)  # 4 v-nodes, 2 features

# Create class_values mapping
class_values = {0: 0.0, 1: 1.0, 2: 2.0}

ARGS = {
    'ind': (0, 0),  # Start from u-node 0, v-node 0
    'Arow': Arow,
    'Acol': Acol,
    'h': 1,  # hop distance
    'sample_ratio': 1.0,
    'max_nodes_per_hop': None,
    'u_features': u_features,
    'v_features': v_features,
    'class_values': class_values,
    'y': 1,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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
count_all_paths_with_mp Not yet run
unverified (harvested, no recorded run)
hwwang55/PathCon/src/utils.py
code served (permissive licence) · get_code("018bf3f65be492cc")
How it ran

no invocation recorded.

load_data Not yet run
unverified (harvested, no recorded run)
muhanzhang/IGMC/data_utils.py
code served (permissive licence) · get_code("3b7d48d27d3a3427")
How it ran

no invocation recorded.

read_relations Not yet run
unverified (harvested, no recorded run)
hwwang55/PathCon/src/data_loader.py
code served (permissive licence) · get_code("b7118c6bf6b1a03c")
How it ran

no invocation recorded.

read_triplets Not yet run
unverified (harvested, no recorded run)
hwwang55/PathCon/src/data_loader.py
code served (permissive licence) · get_code("6165bfc94471e574")
How it ran

no invocation recorded.

test_once Not yet run
unverified (harvested, no recorded run)
muhanzhang/IGMC/train_eval.py
code served (permissive licence) · get_code("8379839e80eb0191")
How it ran

no invocation recorded.

train_multiple_epochs Not yet run
unverified (harvested, no recorded run)
muhanzhang/IGMC/train_eval.py
code served (permissive licence) · get_code("c0108c667c70657c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 18 samples lifted from this paper's linked repositories, 18 carry a function identity and 0 do not. Of the 18 distinct functions with an identity, 7 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 7 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
num_graphs 9 other papers
1 dated earlier, 8 later
1904.12058 (2019-04)
Inductive Matrix Completion Based on Graph Neural Networks
2010.05563 (2020-10)
Graph Information Bottleneck for Subgraph Recognition
2010.12811 (2020-10)
Graph Information Bottleneck
and 6 more not named here
count_all_paths 1 other paper
1 later
2204.13957 (2022-04)
PIE: a Parameter and Inference Efficient Solution for Large Scale Knowledge Graph Embedding Reasoning
count_all_paths_with_mp 1 other paper
1 later
2204.13957 (2022-04)
PIE: a Parameter and Inference Efficient Solution for Large Scale Knowledge Graph Embedding Reasoning
get_params_for_mp 1 other paper
1 later
2204.13957 (2022-04)
PIE: a Parameter and Inference Efficient Solution for Large Scale Knowledge Graph Embedding Reasoning
load_matlab_file 1 other paper
1 later
2205.05957 (2022-05)
Communicative Subgraph Representation Learning for Multi-Relational Inductive Drug-Gene Interaction Prediction
map_data 1 other paper
1 later
2205.05957 (2022-05)
Communicative Subgraph Representation Learning for Multi-Relational Inductive Drug-Gene Interaction Prediction
read_entities 1 other paper
1 later
2204.13957 (2022-04)
PIE: a Parameter and Inference Efficient Solution for Large Scale Knowledge Graph Embedding Reasoning

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 2026-11-21.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Knowledge graph completion aims to predict missing relations between entities in a knowledge graph. In this work, we propose a relational message passing method for knowledge graph completion. Different from existing embedding-based methods, relational message passing only considers edge features (i.e., relation types) without entity IDs in the knowledge graph, and passes relational messages among edges iteratively to aggregate neighborhood information. Specifically, two kinds of neighborhood topology are modeled for a given entity pair under the relational message passing framework: (1) Relational context, which captures the relation types of edges adjacent to the given entity pair; (2) Relational paths, which characterize the relative position between the given two entities in the knowledge graph. The two message passing modules are combined together for relation prediction. Experimental results on knowledge graph benchmarks as well as our newly proposed dataset show that, our method PathCon outperforms state-of-the-art knowledge graph completion methods by a large margin. PathCon is also shown applicable to inductive settings where entities are not seen in training stage, and it is able to provide interpretable explanations for the predicted results. The code and all datasets are available at https://github.com/hwwang55/PathCon.

For agents

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

get_harvested_code_for_paper("2002.06757")
get_code_for_paper("2002.06757")
have("2002.06757")

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

curl https://syntology.ai/api/ran/2002.06757.json

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

[![Syntology run record](https://syntology.ai/api/ran/2002.06757.svg)](https://syntology.ai/paper/2002.06757)

Connect an agent — have() is free.