SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.14040 · CVPR · 2022

Visual Abductive Reasoning

Yi Yang, Chen Liang, Wenguan Wang, Tianfei Zhou

arXiv · PDF · Open in the Atlas

Code that ran

11 of 21 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (8 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 8 have no contract verdict; 3 failures of our instrument; 10 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/2203.14040.json · badge: /api/ran/2203.14040.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
leonnnop/var marked official in Papers with Code
Papers with Code archive
11 ran · 10 no recorded run
FunctionStatusWhere it lives
BertEmbeddingsWithSentence Ran
ran on a synthesized input · constructed an object; no result was computed
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("62fbcc09b795d0d9")
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, BertEmbeddingsWithSentence: 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)

class Config:
    def __init__(self):
        self.vocab_size = 100
        self.word_vec_size = 4
        self.hidden_size = 4
        self.layer_norm_eps = 1e-12
        self.hidden_dropout_prob = 0.1
        self.max_t_len = 5
        self.max_n_len = 3
        self.max_v_len = 4

config = Config()

# Create input tensors
batch_size = 2
seq_len = 5
hidden_size = 4

input_ids = torch.randint(0, config.vocab_size, (batch_size, seq_len))
token_type_ids = torch.randint(0, 3, (batch_size, seq_len))
video_embeddings = torch.randn(batch_size, 4, hidden_size)
sentence_embeddings = torch.randn(batch_size, seq_len, hidden_size)

ARGS = {
    'config': config,
    'add_postion_embeddings': 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 BertEmbeddingsWithSentence — 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 BertEmbeddingsWithSentence 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
BertLMPredictionHead Ran
ran on a synthesized input · constructed an object; no result was computed
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("8dcc67f7cdd708fe")
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, BertLMPredictionHead: 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)

def gelu(x):
    """Implementation of the gelu activation function."""
    return x * 0.5 * (1.0 + torch.erf(x / math.sqrt(2.0)))

class BertLayerNorm(nn.Module):
    def __init__(self, hidden_size, eps=1e-12):
        super(BertLayerNorm, self).__init__()
        self.weight = nn.Parameter(torch.ones(hidden_size))
        self.bias = nn.Parameter(torch.zeros(hidden_size))
        self.variance_epsilon = eps

    def forward(self, x):
        u = x.mean(-1, keepdim=True)
        s = (x - u).pow(2).mean(-1, keepdim=True)
        x = (x - u) / torch.sqrt(s + self.variance_epsilon)
        return self.weight * x + self.bias

class BertPredictionHeadTransform(nn.Module):
    def __init__(self, config):
        super(BertPredictionHeadTransform, self).__init__()
        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
        self.transform_act_fn = gelu
        self.LayerNorm = BertLayerNorm(config.hidden_size, eps=config.layer_norm_eps)

    def forward(self, hidden_states):
        hidden_states = self.dense(hidden_states)
        hidden_states = self.transform_act_fn(hidden_states)
        hidden_states = self.LayerNorm(hidden_states)
        return hidden_states

class BertLMPredictionHead(nn.Module):
    def __init__(self, config, bert_model_embedding_weights=None):
        super(BertLMPredictionHead, self).__init__()
        self.transform = BertPredictionHeadTransform(config)
        if config.share_wd_cls_weight:
            assert bert_model_embedding_weights is not None, 'bert_model_embedding_weights should not be None when setting --share_wd_cls_weight flag to be true'
            assert config.hidden_size == bert_model_embedding_weights.size(1), 'hidden size has be the same as word embedding size when sharing word embedding weight and classifier weight'
            self.decoder = nn.Linear(bert_model_embedding_weights.size(1), bert_model_embedding_weights.size(0), bias=False)
            self.decoder.weight = bert_model_embedding_weights
        else:
            self.decoder = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
        self.bias = nn.Parameter(torch.zeros(config.vocab_size))

    def forward(self, hidden_states):
        hidden_states = self.transform(hidden_states)
        hidden_states = self.decoder(hidden_states) + self.bias
        return hidden_states

class Config:
    def __init__(self, hidden_size, vocab_size, layer_norm_eps, share_wd_cls_weight):
        self.hidden_size = hidden_size
        self.vocab_size = vocab_size
        self.layer_norm_eps = layer_norm_eps
        self.share_wd_cls_weight = share_wd_cls_weight

config = Config(
    hidden_size=4,
    vocab_size=8,
    layer_norm_eps=1e-12,
    share_wd_cls_weight=False
)

ARGS = {
    'config': config,
    'bert_model_embedding_weights': None
}
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 BertLMPredictionHead — 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 BertLMPredictionHead 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
BertPredictionHeadTransform Ran
ran on a synthesized input · constructed an object; no result was computed
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("af63ac23a3503cdf")
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, BertPredictionHeadTransform: 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)

class Config:
    def __init__(self):
        self.hidden_size = 4
        self.layer_norm_eps = 1e-12

ARGS = {
    'config': Config()
}
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 BertPredictionHeadTransform — 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 BertPredictionHeadTransform 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
BertSelfAttention Ran
ran on a synthesized input · constructed an object; no result was computed
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("1ad7e9bae41c84c2")
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, BertSelfAttention: 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 Config:
    def __init__(self):
        self.hidden_size = 8
        self.num_attention_heads = 2
        self.attention_probs_dropout_prob = 0.1
        self.enable_relative = True
        self.max_n_len = 4

config = Config()

ARGS = {
    'config': config
}
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 BertSelfAttention — 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 BertSelfAttention 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
BertSelfAttentionSen Ran
ran on a synthesized input · constructed an object; no result was computed
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("2a92c02716662f19")
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, BertSelfAttentionSen: 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 Config:
    def __init__(self):
        self.hidden_size = 8
        self.num_attention_heads = 2
        self.attention_probs_dropout_prob = 0.1
        self.enable_relative = True
        self.conf_bucket_size = 5

config = Config()

ARGS = {
    'config': config
}
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 BertSelfAttentionSen — 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 BertSelfAttentionSen 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
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("e4821f9d2b6f5330")
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)

# PositionEncoding only takes n_filters and max_len as constructor parameters
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
RelativeTemEmbedding Ran
ran on a synthesized input · constructed an object; no result was computed
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("328e8c6cf27e19c8")
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, RelativeTemEmbedding: 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 config class with required attributes
class MinimalConfig:
    def __init__(self):
        self.hidden_size = 16  # Must be divisible by num_attention_heads
        self.num_attention_heads = 4

config = MinimalConfig()

# Construct the arguments dictionary
ARGS = {
    'config': config,
    'in_lin': False,
    'on_value': False,
    'relative_mode': 'bias',
    'relative_directional': True,
    'relative_multihead': False,
    'num_buckets': 10
}
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 RelativeTemEmbedding — 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 RelativeTemEmbedding 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
TextEmbedding Ran
ran on a synthesized input · constructed an object; no result was computed
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("d3d0c34b93dc9fec")
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, TextEmbedding: 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 Config:
    def __init__(self):
        self.vocab_size = 100
        self.word_vec_size = 4
        self.hidden_size = 8
        self.hidden_dropout_prob = 0.1
        self.layer_norm_eps = 1e-12

config = Config()

ARGS = {
    'config': config
}
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 TextEmbedding — 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 TextEmbedding 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
_concat_list_of_tensors Ran
ran; our contract draft was wrong, not the code
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("9fe227d9a54173f6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function takes a list of lists of tensors
# Each inner list contains tensors that will be stacked
# The tensors have shape (batch_size, *feature_dims)
# After stacking, they're reshaped to (-1, *feature_dims)

# Create a list of lists of tensors
# Outer list represents different parameter groups
# Inner lists contain tensors with batch dimension

inner_list_1 = [
    torch.randn(2, 4),  # batch_size=2, feature_dim=4
    torch.randn(2, 4),  # batch_size=2, feature_dim=4
]

inner_list_2 = [
    torch.randn(3, 5),  # batch_size=3, feature_dim=5
    torch.randn(3, 5),  # batch_size=3, feature_dim=5
]

_list = [inner_list_1, inner_list_2]

ARGS = {
    "_list": _list
}
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 list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
make_pad_shifted_mask Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("ee2eb9135dec6e53")
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 3, dtype family float, shape (2, 6, 6) 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 a simple input_mask tensor
# Shape: (batch_size=2, seq_len=6)
# seq_len = max_v_len + max_t_len + sentence_len = 2 + 3 + 1 = 6
input_mask = torch.ones(2, 6)

# Set some padding to 0 to demonstrate masking
input_mask[0, 5] = 0  # pad the last position for first sample
input_mask[1, 4:] = 0  # pad last two positions for second sample

max_v_len = 2
max_t_len = 3
sentence_len = 1

ARGS = {
    "input_mask": input_mask,
    "max_v_len": max_v_len,
    "max_t_len": max_t_len,
    "sentence_len": sentence_len,
}
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 3, dtype family float, shape (2, 6, 6) — shape (2, 6, 6) 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
make_shifted_mask Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("ed3b9085302e4898")
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 3, dtype family float, shape (2, 6, 6) 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 input_mask: (N, L) where N is batch size and L is sequence length
# L = max_v_len + max_t_len + sentence_len
bsz = 2
max_v_len = 2
max_t_len = 3
sentence_len = 1
seq_len = max_v_len + max_t_len + sentence_len  # 2 + 3 + 1 = 6

# input_mask should be (bsz, seq_len) with 1s for valid, 0s for padding
input_mask = torch.ones(bsz, seq_len)

ARGS = {
    "input_mask": input_mask,
    "max_v_len": max_v_len,
    "max_t_len": max_t_len,
    "sentence_len": sentence_len,
}
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 3, dtype family float, shape (2, 6, 6) — shape (2, 6, 6) 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
BertAttention Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("8a110cded88738a8")
How it ran

no invocation recorded.

BertEmbeddingsWithVideo Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("860f9614a5fe7da7")
How it ran

no invocation recorded.

BertEncoder Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("2fbca415e4e6bf08")
How it ran

no invocation recorded.

BertEncoderSen Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("5aeb53ac7ae0499a")
How it ran

no invocation recorded.

BertEncoderUntied Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("3d9f583a9ae50be3")
How it ran

no invocation recorded.

BertLayer Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("5a8e31314ed696b6")
How it ran

no invocation recorded.

BertLayerSen Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("a677dcf23ed7fc33")
How it ran

no invocation recorded.

BertLayerUntied Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("4d48738bd0f321b8")
How it ran

no invocation recorded.

Reasoner Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("c9602cd24ba31174")
How it ran

no invocation recorded.

VideoEncodingTrans Not yet run
unverified (harvested, no recorded run)
leonnnop/var/src/model/mainmodel.py
code served (permissive licence) · get_code("0efbd684f43cf4f6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 21 samples lifted from this paper's linked repositories, 21 carry a function identity and 0 do not. Of the 21 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 19 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
BertLMPredictionHead
brick of this paper's assembly: helper
2 other papers
1 dated earlier, 1 later · 1 only as an assembly brick
2005.05402 (2020-05)
MART: Memory-Augmented Recurrent Transformer for Coherent Video Paragraph Captioning
2603.05969 (2026-03)
Imagine How To Change: Explicit Procedure Modeling for Change Captioning
BertPredictionHeadTransform
brick of this paper's assembly: helper
2 other papers
1 dated earlier, 1 later · 1 only as an assembly brick
2005.05402 (2020-05)
MART: Memory-Augmented Recurrent Transformer for Coherent Video Paragraph Captioning
2603.05969 (2026-03)
Imagine How To Change: Explicit Procedure Modeling for Change Captioning

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

Abductive reasoning seeks the likeliest possible explanation for partial observations. Although abduction is frequently employed in human daily reasoning, it is rarely explored in computer vision literature. In this paper, we propose a new task and dataset, Visual Abductive Reasoning (VAR), for examining abductive reasoning ability of machine intelligence in everyday visual situations. Given an incomplete set of visual events, AI systems are required to not only describe what is observed, but also infer the hypothesis that can best explain the visual premise. Based on our large-scale VAR dataset, we devise a strong baseline model, REASONER (causal-and-cascaded reasoning Transformer). First, to capture the causal structure of the observations, a contextualized directional position embedding strategy is adopted in the encoder, that yields discriminative representations for the premise and hypothesis. Then, multiple decoders are cascaded to generate and progressively refine the premise and hypothesis sentences. The prediction scores of the sentences are used to guide cross-sentence information flow in the cascaded reasoning procedure. Our VAR benchmarking results show that REASONER surpasses many famous video-language models, while still being far behind human performance. This work is expected to foster future efforts in the reasoning-beyond-observation paradigm.

For agents

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

get_harvested_code_for_paper("2203.14040")
get_code_for_paper("2203.14040")
have("2203.14040")

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

curl https://syntology.ai/api/ran/2203.14040.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.14040.svg)](https://syntology.ai/paper/2203.14040)

Connect an agent — have() is free.