SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1809.10341 · 2018

Deep Graph Infomax

arXiv · PDF · Open in the Atlas

Code that ran

26 of 35 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (11 of them constructed an object rather than computing a result; 2 honoured the contract we drafted, 1 violated it and 19 have no contract verdict; 4 failures of our instrument; 9 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/1809.10341.json · badge: /api/ran/1809.10341.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
PetarV-/DGI canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
wwwyyun/HDGI reimplementation (classifier label)
classifier basis: classifier_2025
5 ran · 1 no recorded run
SpaceLearner/DisenGraphRep reimplementation (classifier label)
classifier basis: classifier_2025
4 ran · 4 no recorded run
Sam131112/Deep-Graph-Infomax-Implementation reimplementation (classifier label)
classifier basis: classifier_2025
4 ran · 0 no recorded run
Sean-Bin-Yang/Path-InfoMax extension (classifier label)
classifier basis: classifier_2025
3 ran · 1 no recorded run
ngminhhieu/DGI community (classifier label)
classifier basis not recorded on the link
3 ran · 0 no recorded run
dfdazac/dgi reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
zpeng27/GMI unrelated (classifier label)
classifier basis: classifier_2025
1 ran · 3 no recorded run
copy not recorded community
not in the paper's declared repositories
1 ran · 0 no recorded run
FunctionStatusWhere it lives
Assign Ran
ran on a synthesized input · constructed an object; no result was computed
wwwyyun/HDGI/models/dgi.py
code served (permissive licence) · get_code("0e9cd1335cbb2931")
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 an instance of its own class, Assign: the run constructed the object and did not call it, so no result was computed. 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)

# Define the Assign module
class Assign(nn.Module):
    def __init__(self, k, in_dim, device, hard):
        super(Assign, self).__init__()
        self.k = k
        self.in_dim = in_dim
        self.community_embed = nn.Parameter(torch.zeros(self.k, self.in_dim))
        self.reset_parameters()
        self.hard = hard
        self.device = device

    def reset_parameters(self):
        torch.nn.init.xavier_uniform_(self.community_embed)

    def communityCal(self, node_embed):
        node_embed = torch.squeeze(node_embed)
        scores = torch.norm(torch.unsqueeze(node_embed, 1) - torch.unsqueeze(self.community_embed, 0), dim=2)
        if self.training:
            assignmat = torch.nn.functional.gumbel_softmax(logits=scores, tau=1.0, hard=self.hard)
        else:
            ind = scores.argmax(dim=-1).reshape(scores.shape[0], 1)
            assignmat = torch.zeros(scores.shape).to(self.device).scatter_(1, ind, 1.0)
        return (self.community_embed, assignmat)

    def forward(self, seq, msk, training):
        self.training = training
        if msk is None:
            return self.communityCal(seq)
        else:
            msk = torch.unsqueeze(msk, -1)
            centroids = self.community_embed
            return torch.sum(centroids * msk, 1) / torch.sum(msk)

device = torch.device('cpu')
k = 3
in_dim = 4

# Create test inputs
seq = torch.randn(2, in_dim)
msk = None
training = True

ARGS = {
    'k': k,
    'in_dim': in_dim,
    'device': device,
    'hard': 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 Assign — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Assign 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
AvgReadout Ran
ran on a synthesized input
PetarV-/DGI/models/dgi.py
code served (permissive licence) · get_code("7f96e512dd0bcc96")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 2, dtype family float, shape (2, 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 torch
import torch.nn as nn

INIT_ARGS = {}

ARGS = {
    'seq': torch.randn(2, 4, 3),
    'msk': torch.ones(2, 4, dtype=torch.bool)
}

if __name__ == '__main__':
    model = nn.Module()
    
    class AvgReadout(nn.Module):
        def __init__(self):
            super(AvgReadout, self).__init__()

        def forward(self, seq, msk):
            if msk is None:
                return torch.mean(seq, 1)
            else:
                msk = torch.unsqueeze(msk, -1)
                return torch.sum(seq * msk, 1) / torch.sum(msk)
    
    readout = AvgReadout()
    output = readout(**ARGS)
    print("Output shape:", output.shape)
    print("Output:", output)
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 3) — shape (2, 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
Bernoulli Ran
ran on a synthesized input · constructed an object; no result was computed
SpaceLearner/DisenGraphRep/models/infomax.py
code served (permissive licence) · get_code("37bd33520c0762ac")
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 an instance of its own class, Bernoulli: the run constructed the object and did not call it, so no result was computed. 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.autograd import Variable

torch.manual_seed(0)

class Bernoulli(nn.Module):
    """Samples from a Bernoulli distribution where the probability is given
    by the sigmoid of the given parameter.
    """

    def __init__(self, p=0.5, stgradient=False):
        super(Bernoulli, self).__init__()
        p = torch.Tensor([p])
        eps = 1e-08
        self.p = Variable(torch.log(p / (1 - p) + eps))
        self.stgradient = stgradient

    def _check_inputs(self, size, ps):
        if size is None and ps is None:
            raise ValueError('Either one of size or params should be provided.')
        elif size is not None and ps is not None:
            if ps.ndimension() > len(size):
                return ps.squeeze(-1).expand(size)
            else:
                return ps.expand(size)
        elif size is not None:
            return self.p.expand(size)
        elif ps is not None:
            return ps
        else:
            raise ValueError('Given invalid inputs: size={}, ps={})'.format(size, ps))

    def _sample_logistic(self, size):
        eps = 1e-08
        u = Variable(torch.rand(size))
        l = torch.log(u + eps) - torch.log(1 - u + eps)
        return l

    def sample(self, size=None, params=None):
        import torch.nn.functional as F
        from torch.autograd import Function
        
        class STHeaviside(Function):
            @staticmethod
            def forward(ctx, x):
                y = torch.zeros(x.size()).type_as(x)
                y[x >= 0] = 1
                return y

            @staticmethod
            def backward(ctx, grad_output):
                return grad_output
        
        eps = 1e-08
        presigm_ps = self._check_inputs(size, params)
        logp = F.logsigmoid(presigm_ps)
        logq = F.logsigmoid(-presigm_ps)
        l = self._sample_logistic(logp.size()).type_as(presigm_ps)
        z = logp - logq + l
        b = STHeaviside.apply(z)
        return b if self.stgradient else b.detach()

    def log_density(self, sample, params=None):
        import torch.nn.functional as F
        eps = 1e-08
        presigm_ps = self._check_inputs(sample.size(), params).type_as(sample)
        p = (F.sigmoid(presigm_ps) + eps) * (1 - 2 * eps)
        logp = sample * torch.log(p + eps) + (1 - sample) * torch.log(1 - p + eps)
        return logp

    def get_params(self):
        return self.p

    @property
    def nparams(self):
        return 1

    @property
    def ndim(self):
        return 1

    @property
    def is_reparameterizable(self):
        return self.stgradient

    def __repr__(self):
        import torch.nn.functional as F
        tmpstr = self.__class__.__name__ + ' ({:.3f})'.format(F.sigmoid(self.p.data)[0])
        return tmpstr


ARGS = {
    'p': 0.5,
    'stgradient': 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 Bernoulli — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Bernoulli 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
DGI Ran
ran on a synthesized input · constructed an object; no result was computed
wwwyyun/HDGI/models/dgi.py
code served (permissive licence) · get_code("d7d2117d73a0c03d")
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 an instance of its own class, DGI: the run constructed the object and did not call it, so no result was computed. 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
import torch.nn.functional as F

torch.manual_seed(0)

# Small stand-in values for the DGI model
n_in = 4          # input feature dimension
n_h = 5           # hidden dimension
activation = 'prelu'  # activation function
k = 3             # number of communities
device = 'cpu'    # CPU only
hard = False      # hard assignment
batch_size = 2    # batch size

# Create tiny test tensors
# seq1, seq2: (batch, num_nodes, features)
seq1 = torch.randn(1, 2, n_in)
seq2 = torch.randn(1, 2, n_in)

# adj: adjacency matrix (batch, num_nodes, num_nodes)
adj = torch.randn(1, 2, 2)

# sparse: boolean flag for sparse operations
sparse = False

# msk: optional mask (batch, num_nodes) or None
msk = None

# samp_bias1, samp_bias2: bias tensors for discriminator
# Based on the code, these should be compatible with sc_1 and sc_2
# sc_1 has shape from self.f_k(h_pl, c_x) which outputs (batch, num_nodes) approximately
samp_bias1 = torch.randn(1, 2)
samp_bias2 = torch.randn(1, 2)

ARGS = {
    'n_in': n_in,
    'n_h': n_h,
    'activation': activation,
    'k': k,
    'device': device,
    'hard': hard,
    'batch_size': batch_size,
}
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 DGI — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a DGI 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
DGI Ran
ran on a synthesized input
PetarV-/DGI/models/dgi.py
code served (permissive licence) · get_code("58c5f9e1c1f1d371")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 2, 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 torch.nn as nn

# Import the classes
from torch.nn import Module, Linear, Bilinear, PReLU, Sigmoid, Parameter

# Define INIT_ARGS for DGI.__init__(n_in, n_h, activation)
INIT_ARGS = {
    'n_in': 4,
    'n_h': 8,
    'activation': 'prelu'
}

# Define ARGS for DGI.forward(seq1, seq2, adj, sparse, msk, samp_bias1, samp_bias2)
ARGS = {
    'seq1': torch.randn(2, 3, 4),  # batch=2, nodes=3, features=4
    'seq2': torch.randn(2, 3, 4),  # batch=2, nodes=3, features=4
    'adj': torch.randn(2, 3, 3),   # batch=2, adjacency 3x3
    'sparse': False,
    'msk': torch.ones(2, 3),       # batch=2, nodes=3
    'samp_bias1': None,
    'samp_bias2': None
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
DeepGraphInfomax Ran
ran on a synthesized input · constructed an object; no result was computed
Sam131112/Deep-Graph-Infomax-Implementation/deepgraphinfomax.py
pointer only (licence: NONE) · get_code("18a11990837977e4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sam131112/Deep-Graph-Infomax-Implementation 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.

Discriminator Ran
ran on a synthesized input · constructed an object; no result was computed
Sam131112/Deep-Graph-Infomax-Implementation/deepgraphinfomax.py
pointer only (licence: NONE) · get_code("8ac4c2aa6c924a58")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sam131112/Deep-Graph-Infomax-Implementation 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.

Discriminator Ran
ran on a synthesized input · constructed an object; no result was computed
Sean-Bin-Yang/Path-InfoMax/models/pim.py
pointer only (licence: NONE) · get_code("3e0b949d70c76ab4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sean-Bin-Yang/Path-InfoMax 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.

Discriminator Ran
ran on a synthesized input · constructed an object; no result was computed
PetarV-/DGI/models/dgi.py
code served (permissive licence) · get_code("1702410d0868c8a5")
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 an instance of its own class, Discriminator: the run constructed the object and did not call it, so no result was computed. 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 Discriminator(nn.Module):
    def __init__(self, n_h):
        super(Discriminator, self).__init__()
        self.f_k = nn.Bilinear(n_h, n_h, 1)
        for m in self.modules():
            self.weights_init(m)

    def weights_init(self, m):
        if isinstance(m, nn.Bilinear):
            torch.nn.init.xavier_uniform_(m.weight.data)
            if m.bias is not None:
                m.bias.data.fill_(0.0)

    def forward(self, c, h_pl, h_mi, s_bias1=None, s_bias2=None):
        c_x = torch.unsqueeze(c, 1)
        c_x = c_x.expand_as(h_pl)
        sc_1 = torch.squeeze(self.f_k(h_pl, c_x), 2)
        sc_2 = torch.squeeze(self.f_k(h_mi, c_x), 2)
        if s_bias1 is not None:
            sc_1 += s_bias1
        if s_bias2 is not None:
            sc_2 += s_bias2
        logits = torch.cat((sc_1, sc_2), 1)
        return logits

n_h = 4
discriminator = Discriminator(n_h)

batch_size = 2
c = torch.randn(batch_size, n_h)
h_pl = torch.randn(batch_size, 3, n_h)
h_mi = torch.randn(batch_size, 3, n_h)
s_bias1 = torch.randn(batch_size, 3)
s_bias2 = torch.randn(batch_size, 3)

ARGS = {
    "n_h": n_h,
}
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 Discriminator — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Discriminator 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
EncoderLSTM Ran
ran on a synthesized input · constructed an object; no result was computed
Sean-Bin-Yang/Path-InfoMax/models/pim.py
pointer only (licence: NONE) · get_code("9a35f79723ddacf3")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sean-Bin-Yang/Path-InfoMax 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.

GCN Ran
ran on a synthesized input · constructed an object; no result was computed
wwwyyun/HDGI/models/dgi.py
code served (permissive licence) · get_code("5fd25f0307b96798")
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 an instance of its own class, GCN: the run constructed the object and did not call it, so no result was computed. 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 GCN(nn.Module):

    def __init__(self, in_ft, out_ft, act, bias=True):
        super(GCN, self).__init__()
        self.fc = nn.Linear(in_ft, out_ft, bias=False)
        self.act = nn.PReLU() if act == 'prelu' else act
        if bias:
            self.bias = nn.Parameter(torch.FloatTensor(out_ft))
            self.bias.data.fill_(0.0)
        else:
            self.register_parameter('bias', None)
        for m in self.modules():
            self.weights_init(m)

    def weights_init(self, m):
        if isinstance(m, nn.Linear):
            torch.nn.init.xavier_uniform_(m.weight.data)
            if m.bias is not None:
                m.bias.data.fill_(0.0)

    def forward(self, seq, adj, sparse=False):
        seq_fts = self.fc(seq)
        if sparse:
            out = torch.unsqueeze(torch.spmm(adj, torch.squeeze(seq_fts, 0)), 0)
        else:
            out = torch.bmm(adj, seq_fts)
        if self.bias is not None:
            out += self.bias
        return self.act(out)


in_ft = 5
out_ft = 4
act = nn.ReLU()
bias = True

seq = torch.randn(2, 3, 5)
adj = torch.randn(2, 3, 3)
sparse = False

ARGS = {
    'in_ft': in_ft,
    'out_ft': out_ft,
    'act': act,
    'bias': bias
}
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 GCN — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a GCN 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
GCN Ran
ran on a synthesized input
zpeng27/GMI/models/gmi.py
code served (permissive licence) · get_code("1efc738df1839678")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) 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
import numpy as np

# INIT_ARGS for GCN.__init__(in_ft, out_ft, act, bias=True)
INIT_ARGS = {
    'in_ft': 5,
    'out_ft': 7,
    'act': 'prelu',
    'bias': True
}

# Create a sparse adjacency matrix for the forward pass
# adj should be a sparse tensor representing graph connectivity
num_nodes = 6
indices = torch.LongTensor([[0, 1, 1, 2, 2, 3, 3, 4, 4, 5],
                             [1, 0, 2, 1, 3, 2, 4, 3, 5, 4]])
values = torch.FloatTensor([1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0])
adj = torch.sparse_coo_tensor(indices, values, (num_nodes, num_nodes))

# seq should be node features: (batch=1, num_nodes, in_ft)
seq = torch.randn(1, num_nodes, INIT_ARGS['in_ft'])

# ARGS for GCN.forward(seq, adj)
ARGS = {
    'seq': seq,
    'adj': adj
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
GraphConvolutionLayer Ran
ran on a synthesized input
Sam131112/Deep-Graph-Infomax-Implementation/deepgraphinfomax.py
pointer only (licence: NONE) · get_code("3410df5150572826")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sam131112/Deep-Graph-Infomax-Implementation 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.

Normal Ran
ran on a synthesized input · constructed an object; no result was computed
SpaceLearner/DisenGraphRep/models/infomax.py
code served (permissive licence) · get_code("7af5f07d76374e95")
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 an instance of its own class, Normal: the run constructed the object and did not call it, so no result was computed. 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
import numpy as np
import math
from torch.autograd import Variable

torch.manual_seed(0)

# Initialize the Normal module
normal_dist = nn.Module()

# The __init__ parameters for Normal class are:
# - mu (default 0)
# - sigma (default 1)

ARGS = {
    "mu": 0,
    "sigma": 1
}
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 Normal — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Normal 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
Readout Ran
ran on a synthesized input
Sam131112/Deep-Graph-Infomax-Implementation/deepgraphinfomax.py
pointer only (licence: NONE) · get_code("5c44b5cacba28aaa")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sam131112/Deep-Graph-Infomax-Implementation 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.

Readout Ran
ran on a synthesized input
Sean-Bin-Yang/Path-InfoMax/models/pim.py
pointer only (licence: NONE) · get_code("ebd1fd43842156d6")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sean-Bin-Yang/Path-InfoMax 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.

STHeaviside Ran
ran on a synthesized input · constructed an object; no result was computed
SpaceLearner/DisenGraphRep/models/infomax.py
code served (permissive licence) · get_code("72705e2c2924fe8c")
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 an instance of its own class, STHeaviside: the run constructed the object and did not call it, so no result was computed. 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
from torch.autograd import Function

torch.manual_seed(0)

class STHeaviside(Function):

    @staticmethod
    def forward(ctx, x):
        y = torch.zeros(x.size()).type_as(x)
        y[x >= 0] = 1
        return y

    @staticmethod
    def backward(ctx, grad_output):
        return grad_output


# Create a small input tensor
x = torch.randn(2, 4, requires_grad=True)

ARGS = {
    'x': x
}
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 STHeaviside — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a STHeaviside 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
entropy Ran
ran, honoured the contract we drafted
wwwyyun/HDGI/execute.py
code served (permissive licence) · get_code("f5bc6debaeb77061")
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 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

torch.manual_seed(0)

# The function entropy takes a single parameter x which should be a tensor
# It performs operations: sum, div, log, and sum on the tensor
# So x should be a 1D or multi-dimensional tensor with numeric values

x = torch.tensor([1.0, 2.0, 3.0, 4.0])

ARGS = {
    "x": x
}
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 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
idx2onehot Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ngminhhieu/DGI/utils/dgi/process.py
pointer only (licence: MIT); the stored copy get_code serves, from timbmg/VAE-CVAE-MNIST, is pointer only (licence: NONE) · get_code("e0eb450eeb67bbe7")
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 timbmg/VAE-CVAE-MNIST, 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.

load_data Ran
ran; our contract draft was wrong, not the code
dfdazac/dgi/gcn/utils.py
code served (permissive licence) · get_code("36d8bb4858e7bb51")
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 numpy as np
import pickle as pkl
import networkx as nx
import scipy.sparse as sp
import tempfile
import os

# Create a temporary directory for mock data files
temp_dir = tempfile.mkdtemp()
data_dir = os.path.join(temp_dir, 'data')
os.makedirs(data_dir, exist_ok=True)

# Create mock dataset files
dataset_str = 'citeseer'

# Number of nodes and features
n_nodes = 100
n_features = 10
n_classes = 6
n_train = 20
n_val = 500
n_test = 30

# Create sparse feature matrices
x = sp.random(n_train, n_features, density=0.5, format='csr')
tx = sp.random(n_test, n_features, density=0.5, format='csr')
allx = sp.vstack([x, sp.random(n_val, n_features, density=0.5, format='csr')], format='csr')

# Create one-hot labels
y = np.eye(n_classes)[np.random.randint(0, n_classes, n_train)]
ty = np.eye(n_classes)[np.random.randint(0, n_classes, n_test)]
ally = np.vstack([y, np.eye(n_classes)[np.random.randint(0, n_classes, n_val)]])

# Create a graph as dict of lists
graph = {i: list(np.random.choice(n_nodes, size=np.random.randint(1, 5), replace=False)) 
         for i in range(n_nodes)}

# Save pickle files
with open(os.path.join(data_dir, f'ind.{dataset_str}.x'), 'wb') as f:
    pkl.dump(x, f)
with open(os.path.join(data_dir, f'ind.{dataset_str}.y'), 'wb') as f:
    pkl.dump(y, f)
with open(os.path.join(data_dir, f'ind.{dataset_str}.tx'), 'wb') as f:
    pkl.dump(tx, f)
with open(os.path.join(data_dir, f'ind.{dataset_str}.ty'), 'wb') as f:
    pkl.dump(ty, f)
with open(os.path.join(data_dir, f'ind.{dataset_str}.allx'), 'wb') as f:
    pkl.dump(allx, f)
with open(os.path.join(data_dir, f'ind.{dataset_str}.ally'), 'wb') as f:
    pkl.dump(ally, f)
with open(os.path.join(data_dir, f'ind.{dataset_str}.graph'), 'wb') as f:
    pkl.dump(graph, f)

# Create test index file
test_idx_reorder = np.arange(n_train, n_train + n_test)
with open(os.path.join(data_dir, f'ind.{dataset_str}.test.index'), 'w') as f:
    for idx in test_idx_reorder:
        f.write(f'{idx}\n')

# Change to temp directory so relative paths work
original_dir = os.getcwd()
os.chdir(temp_dir)

ARGS = {
    'dataset_str': dataset_str
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
networkx, 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
logsumexp Ran
ran; our contract draft was wrong, not the code
SpaceLearner/DisenGraphRep/models/infomax.py
code served (permissive licence) · get_code("ef2e8969dc4b82fe")
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 1, dtype family float, shape (2,) 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 math
import torch
from numbers import Number

torch.manual_seed(0)

# Test case 1: 2D tensor with dim specified
value_2d = torch.randn(2, 4)

# Test case 2: 1D tensor without dim (scalar result)
value_1d = torch.randn(5)

# Test case 3: 3D tensor with keepdim=True
value_3d = torch.randn(2, 3, 4)

# We'll test with the 2D tensor, dim=1, and keepdim=False as the main case
ARGS = {
    "value": value_2d,
    "dim": 1,
    "keepdim": 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 Tensor, ndim 1, dtype family float, shape (2,) — shape (2,) 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
parse_index_file Ran
ran, honoured the contract we drafted
this paper's copy was not recorded; identical code first harvested from DSL-Lab/Specformer
pointer only; the stored copy get_code serves, from DSL-Lab/Specformer, is pointer only (licence: GPL-3.0) · get_code("c1d6392f89c5e495")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

parse_skipgram Ran
ran on a synthesized input
ngminhhieu/DGI/utils/dgi/process.py
code served (permissive licence) · get_code("47c9a03fbd1720a3")
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 (3, 4) 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 tempfile
import os

# Create a temporary file with valid skipgram format
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_fname = temp_file.name

# Write valid skipgram format:
# First line: nb_nodes nb_features
# Then for each node: node_id (1-indexed) followed by feature values
content = """3 4
1 0.1 0.2 0.3 0.4
2 0.5 0.6 0.7 0.8
3 0.9 1.0 1.1 1.2
"""

temp_file.write(content)
temp_file.close()

ARGS = {
    "fname": temp_fname
}
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 (3, 4) — shape (3, 4) 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
process_tu Ran
ran on a synthesized input
ngminhhieu/DGI/utils/dgi/process.py
code served (permissive licence) · get_code("10796842869164d2")
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 torch
import numpy as np
import scipy.sparse as sp

torch.manual_seed(0)

class MockGraph:
    def __init__(self, x, y, edge_index):
        self.x = x
        self.y = y
        self.edge_index = edge_index

class MockDataset:
    def __init__(self, graphs, num_features):
        self.graphs = graphs
        self.num_features = num_features
    
    def __len__(self):
        return len(self.graphs)
    
    def __getitem__(self, idx):
        return self.graphs[idx]

# Create mock graphs with varying sizes
num_features = 4
graph1_nodes = 3
graph2_nodes = 2

# Graph 1: 3 nodes, 4 features
x1 = np.random.randn(graph1_nodes, num_features).astype(np.float32)
y1 = np.array([0])
edge_index1 = np.array([[0, 1, 2], [1, 2, 0]], dtype=np.int64)

# Graph 2: 2 nodes, 4 features
x2 = np.random.randn(graph2_nodes, num_features).astype(np.float32)
y2 = np.array([1])
edge_index2 = np.array([[0, 1], [1, 0]], dtype=np.int64)

graph1 = MockGraph(x1, y1, edge_index1)
graph2 = MockGraph(x2, y2, edge_index2)

# Create dataset with both graphs
data = MockDataset([graph1, graph2], num_features=num_features)

# Maximum number of nodes across all graphs
nb_nodes = 5

ARGS = {
    "data": data,
    "nb_nodes": nb_nodes
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, 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
prune Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
wwwyyun/HDGI/execute.py
code served (permissive licence) · get_code("e853d6332902fb5f")
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 Tensor, ndim 1, dtype family float, shape (4,) 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)

# Construct edge_index: shape (2, num_edges)
# This represents edges in a graph with row indices and column indices
edge_index = torch.tensor([
    [0, 1, 0, 1],  # row indices
    [1, 0, 1, 0]   # col indices
], dtype=torch.long)

# Construct y: node embeddings with shape (num_nodes, embedding_dim)
# We need at least 2 nodes (indices 0 and 1 appear in edge_index)
# Using embedding_dim=4 as specified
y = torch.randn(2, 4)

ARGS = {
    "edge_index": edge_index,
    "y": y
}
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 Tensor, ndim 1, dtype family float, shape (4,) — shape (4,) 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
sample_mask Ran
ran, violated the contract we drafted
dfdazac/dgi/gcn/utils.py
code served (permissive licence) · get_code("b32fd748b4bbe3bf")
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 ndarray, ndim 1, dtype family float, shape (5,) 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

# Analyze the function:
# - idx: used as an index into mask array, should be an integer or array of integers
# - l: used as the length parameter for np.zeros, should be an integer

# Create valid arguments
idx = 2  # Single index into the mask array
l = 5    # Length of the mask array

ARGS = {
    "idx": idx,
    "l": l
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 1, dtype family float, shape (5,) — shape (5,) 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
AvgNeighbor Not yet run
unverified (harvested, no recorded run)
zpeng27/GMI/models/gmi.py
code served (permissive licence) · get_code("27f7cf29d9108df6")
How it ran

no invocation recorded.

DeepGraphInfomax Not yet run
unverified (harvested, no recorded run)
SpaceLearner/DisenGraphRep/models/infomax.py
code served (permissive licence) · get_code("5aa5707ddd7c40ca")
How it ran

no invocation recorded.

Discriminator Not yet run
unverified (harvested, no recorded run)
wwwyyun/HDGI/models/dgi.py
code served (permissive licence) · get_code("690842e491f017bd")
How it ran

no invocation recorded.

Discriminator Not yet run
unverified (harvested, no recorded run)
zpeng27/GMI/models/gmi.py
code served (permissive licence) · get_code("e31d20ab111c4457")
How it ran

no invocation recorded.

GMI Not yet run
unverified (harvested, no recorded run)
zpeng27/GMI/models/gmi.py
code served (permissive licence) · get_code("e1c0b15be5eb9e7f")
How it ran

no invocation recorded.

PIM Not yet run
unverified (harvested, no recorded run)
Sean-Bin-Yang/Path-InfoMax/models/pim.py
pointer only (licence: NONE) · get_code("17418975d54cd5c9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Sean-Bin-Yang/Path-InfoMax 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. No invocation recorded.

logsumexp Not yet run
unverified (harvested, no recorded run)
SpaceLearner/DisenGraphRep/models/dgi.py
code served (permissive licence) · get_code("5b63bfbc3d382b97")
How it ran

no invocation recorded.

reset Not yet run
unverified (harvested, no recorded run)
SpaceLearner/DisenGraphRep/models/infomax.py
code served (permissive licence) · get_code("63491bdac79e3e64")
How it ran

no invocation recorded.

uniform Not yet run
unverified (harvested, no recorded run)
SpaceLearner/DisenGraphRep/models/infomax.py
code served (permissive licence) · get_code("aee2e93dc7075ef3")
How it ran

no invocation recorded.

Also present in other papers' code

Of 35 samples lifted from this paper's linked repositories, 35 carry a function identity and 0 do not. Of the 34 distinct functions with an identity, 11 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 10 later) and 23 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
parse_index_file 53 other papers
6 dated earlier, 47 later · 1 only as an assembly brick
1603.08861 (2016-03)
Revisiting Semi-Supervised Learning with Graph Embeddings
1611.07308 (2016-11)
Variational Graph Auto-Encoders
1710.10568 (2017-10)
Stochastic Training of Graph Convolutional Networks with Variance Reduction
and 50 more not named here
sample_mask 28 other papers
4 dated earlier, 24 later
1603.08861 (2016-03)
Revisiting Semi-Supervised Learning with Graph Embeddings
1710.10568 (2017-10)
Stochastic Training of Graph Convolutional Networks with Variance Reduction
1710.10903 (2017-10)
Graph Attention Networks
and 25 more not named here
logsumexp
brick of this paper's assembly: helper
5 other papers
2 dated earlier, 3 later
1710.11417 (2017-10)
TreeQN and ATreeC: Differentiable Tree-Structured Models for Deep Reinforcement Learning
1802.04942 (2018-02)
Isolating Sources of Disentanglement in Variational Autoencoders
1906.02735 (2019-06)
Residual Flows for Invertible Generative Modeling
and 2 more not named here
load_data 4 other papers
4 later
1902.06667 (2019-02)
Hierarchical Graph Convolutional Networks for Semi-supervised Node Classification
1905.13403 (2019-05)
Deep Bayesian Optimization on Attributed Graphs
2006.11468 (2020-06)
Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
and 1 more not named here
GCN
brick of this paper's assembly: model
2 other papers
2 later · 1 only as an assembly brick
2402.11887 (2024-02)
Generative Semi-supervised Graph Anomaly Detection
2510.02014 (2025-10)
Normality Calibration in Semi-supervised Graph Anomaly Detection
reset
brick of this paper's assembly: helper
2 other papers
2 later · 1 only as an assembly brick
2202.12449 (2022-02)
Directed Graph Auto-Encoders
2305.12074 (2023-05)
DisCo: Distilled Student Models Co-training for Semi-supervised Text Mining
STHeaviside
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2106.05241 (2021-06)
Multi-Facet Clustering Variational Autoencoders
idx2onehot 1 other paper
1 dated earlier
1312.6114 (2013-12)
Auto-Encoding Variational Bayes
parse_skipgram 1 other paper
1 later
2108.09896 (2021-08)
Generative and Contrastive Self-Supervised Learning for Graph Anomaly Detection
process_tu 1 other paper
1 later
2108.09896 (2021-08)
Generative and Contrastive Self-Supervised Learning for Graph Anomaly Detection
uniform
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2005.00791 (2020-05)
KinGDOM: Knowledge-Guided DOMain Adaptation for Sentiment Analysis

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

We present Deep Graph Infomax (DGI), a general approach for learning node representations within graph-structured data in an unsupervised manner. DGI relies on maximizing mutual information between patch representations and corresponding high-level summaries of graphs---both derived using established graph convolutional network architectures. The learnt patch representations summarize subgraphs centered around nodes of interest, and can thus be reused for downstream node-wise learning tasks. In contrast to most prior approaches to unsupervised learning with GCNs, DGI does not rely on random walk objectives, and is readily applicable to both transductive and inductive learning setups. We demonstrate competitive performance on a variety of node classification benchmarks, which at times even exceeds the performance of supervised learning.

For agents

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

get_harvested_code_for_paper("1809.10341")
get_code_for_paper("1809.10341")
have("1809.10341")

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

curl https://syntology.ai/api/ran/1809.10341.json

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

[![Syntology run record](https://syntology.ai/api/ran/1809.10341.svg)](https://syntology.ai/paper/1809.10341)

Connect an agent — have() is free.