SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.10446 · ACL · 2021

Attend What You Need: Motion-Appearance Synergistic Networks for Video Question Answering

Byoung-Tak Zhang, Gi-Cheon Kang, Ahjeong Seo, Joonhan Park

arXiv · PDF · Open in the Atlas

Code that ran

15 of 20 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (14 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 15 have no contract verdict; 0 failures of our instrument; 5 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/2106.10446.json · badge: /api/ran/2106.10446.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
ahjeongseo/MASN-pytorch marked official in Papers with Code
Papers with Code archive
15 ran · 5 no recorded run
FunctionStatusWhere it lives
AttAdj Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("bd681740f90f16fd")
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, AttAdj: 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 math

torch.manual_seed(0)

# Construct the AttAdj module
hidden_size = 4
model = nn.Module()  # placeholder if needed

# Create sample tensors for forward pass testing
batch_size = 2
kv_length = 3
q_length = 5
hidden_size = 4

# k: (batch, kv_length, hidden_size)
k = torch.randn(batch_size, kv_length, hidden_size)

# q: (batch, q_length, hidden_size)
q = torch.randn(batch_size, q_length, hidden_size)

# mask: (batch, q_length, kv_length) - optional
mask = torch.ones(batch_size, q_length, kv_length)

ARGS = {
    "hidden_size": 4,
    "dropout_r": 0.1,
}
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 AttAdj — 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 AttAdj 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
AttFlat Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("759b0914ac77b28a")
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, AttFlat: 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 ARGS dict with all parameters for AttFlat.__init__
ARGS = {
    'hidden_size': 8,
    'flat_mlp_size': 16,
    'flat_glimpses': 3,
    'flat_out_size': 12,
    'dropout_r': 0.1,
}

# Create an instance to verify it works
model = nn.Module()
att_flat = __import__('torch.nn', fromlist=['Module'])

# Test forward pass with minimal tensors
batch_size = 2
seq_len = 4
hidden_size = 8
flat_glimpses = 3

x = torch.randn(batch_size, seq_len, hidden_size)
x_mask = torch.ones(batch_size, seq_len, 1)

att_flat_module = __import__('sys').modules[__name__]
# Note: The actual instantiation and forward pass would happen in the execution context
# We're just providing the argument dictionary as required

ARGS = {
    'hidden_size': 8,
    'flat_mlp_size': 16,
    'flat_glimpses': 3,
    'flat_out_size': 12,
    'dropout_r': 0.1,
}
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 AttFlat — 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 AttFlat 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
BCNet Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("e259237fcad3dc88")
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, BCNet: 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)

ARGS = {
    'v_dim': 4,
    'q_dim': 5,
    'h_dim': 6,
    'h_out': 8,
    'act': 'ReLU',
    'dropout': [0.2, 0.5],
    'k': 3
}
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 BCNet — 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 BCNet 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
BaseRNN Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("dc8f9d8e6494711c")
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, BaseRNN: 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)

ARGS = {
    'input_size': 10,
    'hidden_size': 8,
    'input_dropout_p': 0.1,
    'dropout_p': 0.2,
    'n_layers': 2,
    'rnn_cell': 'lstm'
}
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 BaseRNN — 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 BaseRNN 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
FC Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("05b99b050c8b7491")
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, FC: 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)

ARGS = {
    'in_size': 4,
    'out_size': 3,
    'dropout_r': 0.5,
    'use_relu': True
}
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 FC — 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 FC 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
FCNet Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("63433a241385e836")
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, FCNet: 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.nn.utils.weight_norm import weight_norm

torch.manual_seed(0)

ARGS = {
    'dims': [4, 8, 6, 2],
    'act': 'ReLU',
    'dropout': 0.1
}
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 FCNet — 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 FCNet 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
FFN Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("ccbe74241dafb0f7")
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, FFN: 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)

ARGS = {
    "hidden_size": 64,
    "ff_size": 128,
    "dropout_r": 0.1,
}
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 FFN — 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 FFN 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 · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("b86d984ed27f78a4")
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)

# Define the GCN model parameters
input_size = 4
hidden_size = 8
num_classes = 3
num_layers = 2
dropout = 0.1

# Create the GCN model
model = nn.Module()  # This is not used, but keeping structure clean

# Create sample input tensors
# x: batch_size=2, sequence_length=5, input_size=4
x = torch.randn(2, 5, 4)

# adj: batch_size=2, sequence_length=5, sequence_length=5 (adjacency matrix for each batch)
adj = torch.randn(2, 5, 5)

# Define ARGS dict with all parameters for GCN.__init__
ARGS = {
    'input_size': input_size,
    'hidden_size': hidden_size,
    'num_classes': num_classes,
    'num_layers': num_layers,
    'dropout': dropout
}
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
GraphConvolution Ran
ran on a synthesized input
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("388325a7a33168a2")
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 3, 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
from torch.nn.modules.module import Module

class GraphConvolution(Module):
    """
    Simple GCN layer, similar to https://arxiv.org/abs/1609.02907
    """

    def __init__(self, in_features, out_features):
        super(GraphConvolution, self).__init__()
        self.weight = nn.Linear(in_features, out_features, bias=False)
        self.layer_norm = nn.LayerNorm(out_features, elementwise_affine=False)

    def forward(self, input, adj):
        support = self.weight(input)
        output = torch.bmm(adj, support)
        output = self.layer_norm(output)
        return output


INIT_ARGS = {
    'in_features': 4,
    'out_features': 3,
}

ARGS = {
    'input': torch.randn(2, 5, 4),
    'adj': torch.randn(2, 5, 5),
}

if __name__ == '__main__':
    gcn = GraphConvolution(**INIT_ARGS)
    output = gcn(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Expected shape: (2, 5, 3)")
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 3, 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)
MLP Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("14b57ad597a09b12")
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, MLP: 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 FC(nn.Module):

    def __init__(self, in_size, out_size, dropout_r=0.0, use_relu=True):
        super(FC, self).__init__()
        self.dropout_r = dropout_r
        self.use_relu = use_relu
        self.linear = nn.Linear(in_size, out_size)
        if use_relu:
            self.relu = nn.ReLU(inplace=True)
        if dropout_r > 0:
            self.dropout = nn.Dropout(dropout_r)

    def forward(self, x):
        x = self.linear(x)
        if self.use_relu:
            x = self.relu(x)
        if self.dropout_r > 0:
            x = self.dropout(x)
        return x

class MLP(nn.Module):

    def __init__(self, in_size, mid_size, out_size, dropout_r=0.0, use_relu=True):
        super(MLP, self).__init__()
        self.fc = FC(in_size, mid_size, dropout_r=dropout_r, use_relu=use_relu)
        self.linear = nn.Linear(mid_size, out_size)

    def forward(self, x):
        return self.linear(self.fc(x))

ARGS = {
    'in_size': 4,
    'mid_size': 8,
    'out_size': 3,
    'dropout_r': 0.1,
    'use_relu': True
}

mlp = MLP(**ARGS)
x = torch.randn(2, 4)
output = mlp(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 MLP — 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 MLP 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
PositionEncoding Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("8df81fac54e85c99")
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, PositionEncoding: 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 math

torch.manual_seed(0)

ARGS = {
    'n_filters': 128,
    'max_len': 500
}
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 PositionEncoding — 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 PositionEncoding 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
SentenceEncoderRNN Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("ef071499ee045c12")
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, SentenceEncoderRNN: 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)

# Create a minimal SentenceEncoderRNN with various configurations
# We'll test with bidirectional, multi-layer, and variable length inputs

# Tiny tensor dimensions for testing
batch_size = 2
seq_len = 5
input_dim = 4  # embedding dimension (distinct from hidden_size)
hidden_size = 3  # distinct from input_dim
n_layers = 2
vocab_size = 10

# Create input tensor: (batch, seq_len, input_dim)
input_var = torch.randn(batch_size, seq_len, input_dim)

# Create input lengths for variable length sequences
input_lengths = torch.tensor([5, 3], dtype=torch.long)

# Initial hidden state (optional, can be None)
# For GRU: (num_layers * num_directions, batch, hidden_size)
h_0 = torch.randn(n_layers, batch_size, hidden_size)

ARGS = {
    'vocab_size': vocab_size,
    'hidden_size': hidden_size,
    'input_dropout_p': 0.1,
    'dropout_p': 0.1,
    'n_layers': n_layers,
    'bidirectional': True,
    'rnn_cell': 'gru',
    'variable_lengths': True,
}
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 SentenceEncoderRNN — 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 SentenceEncoderRNN 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
VideoAdjLearner Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("b54092a60a1899c5")
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, VideoAdjLearner: 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)

# Create a VideoAdjLearner instance to test
in_feature_dim = 4
hidden_size = 3
dropout = 0.1
scale = 100

model = nn.Module()

# Prepare input tensors
batch_size = 2
v_len = 5

# v: (b, v_len, d)
v = torch.randn(batch_size, v_len, in_feature_dim)

# v_mask: (b, v_len)
v_mask = torch.ones(batch_size, v_len)

ARGS = {
    "in_feature_dim": in_feature_dim,
    "hidden_size": hidden_size,
    "dropout": dropout,
    "scale": scale
}
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 VideoAdjLearner — 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 VideoAdjLearner 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
WeightDropLinear Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("7c200f265dc848c2")
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, WeightDropLinear: 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.nn import Parameter

torch.manual_seed(0)

ARGS = {
    'in_features': 4,
    'out_features': 3,
    'bias': True,
    'weight_dropout': 0.5,
}
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 WeightDropLinear — 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 WeightDropLinear 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
weight_drop Ran
ran on a synthesized input · constructed an object; no result was computed
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("3d02a8046580e19b")
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, weight_drop: 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.nn import Parameter

torch.manual_seed(0)

# Create a simple module to apply weight_drop to
class SimpleModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.weight = Parameter(torch.randn(4, 3))
        self.bias = Parameter(torch.randn(3))
    
    def forward(self, x):
        return torch.matmul(x, self.weight) + self.bias

module = SimpleModule()

weights = ['weight']

dropout = 0.5

ARGS = {
    'module': module,
    'weights': weights,
    'dropout': dropout
}
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 weight_drop — 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 weight_drop 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
BAN Not yet run
unverified (harvested, no recorded run)
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("1e41b6c5ad0bb279")
How it ran

no invocation recorded.

BiAttention Not yet run
unverified (harvested, no recorded run)
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("84b7ec1fe8c2e0b6")
How it ran

no invocation recorded.

MASN Not yet run
unverified (harvested, no recorded run)
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("119ae32ced71cbed")
How it ran

no invocation recorded.

MotionApprFusion Not yet run
unverified (harvested, no recorded run)
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("f38305bce2e96714")
How it ran

no invocation recorded.

_weight_drop Not yet run
unverified (harvested, no recorded run)
ahjeongseo/MASN-pytorch/model/masn.py
code served (permissive licence) · get_code("73eb643334aa5e32")
How it ran

no invocation recorded.

Also present in other papers' code

Of 20 samples lifted from this paper's linked repositories, 20 carry a function identity and 0 do not. Of the 20 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 2 later) and 18 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
MLP
brick of this paper's assembly: helper
3 other papers
1 dated earlier, 2 later · 2 only as an assembly brick
2105.14517 (2021-05)
GeoQA: A Geometric Question Answering Benchmark Towards Multimodal Numerical Reasoning
2505.18686 (2025-05)
WeakMCN: Multi-task Collaborative Network for Weakly Supervised Referring Expression Comprehension and Segmentation
2508.01984 (2025-08)
IMoRe: Implicit Program-Guided Reasoning for Human Motion Q&A
FC
brick of this paper's assembly: helper
2 other papers
1 dated earlier, 1 later · 2 only as an assembly brick
2105.14517 (2021-05)
GeoQA: A Geometric Question Answering Benchmark Towards Multimodal Numerical Reasoning
2505.18686 (2025-05)
WeakMCN: Multi-task Collaborative Network for Weakly Supervised Referring Expression Comprehension and Segmentation

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

Video Question Answering is a task which requires an AI agent to answer questions grounded in video. This task entails three key challenges: (1) understand the intention of various questions, (2) capturing various elements of the input video (e.g., object, action, causality), and (3) cross-modal grounding between language and vision information. We propose Motion-Appearance Synergistic Networks (MASN), which embed two crossmodal features grounded on motion and appearance information and selectively utilize them depending on the question's intentions. MASN consists of a motion module, an appearance module, and a motion-appearance fusion module. The motion module computes the action-oriented cross-modal joint representations, while the appearance module focuses on the appearance aspect of the input video. Finally, the motion-appearance fusion module takes each output of the motion module and the appearance module as input, and performs question-guided fusion. As a result, MASN achieves new state-of-the-art performance on the TGIF-QA and MSVD-QA datasets. We also conduct qualitative analysis by visualizing the inference results of MASN. The code is available at https://github.com/ ahjeongseo/MASN-pytorch.

For agents

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

get_harvested_code_for_paper("2106.10446")
get_code_for_paper("2106.10446")
have("2106.10446")

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

curl https://syntology.ai/api/ran/2106.10446.json

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

[![Syntology run record](https://syntology.ai/api/ran/2106.10446.svg)](https://syntology.ai/paper/2106.10446)

Connect an agent — have() is free.