SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1611.07810 · 2016

A dataset and exploration of models for understanding video data through fill-in-the-blank question-answering

arXiv · PDF · Open in the Atlas

Code that ran

6 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 2 failures of our instrument; 2 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/1611.07810.json · badge: /api/ran/1611.07810.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
totalgood/viddesc community (classifier label)
classifier basis not recorded on the link
6 ran · 2 no recorded run
FunctionStatusWhere it lives
collate_fn Ran
ran; our contract draft was wrong, not the code
totalgood/viddesc/src/imgdesc/data_loader.py
code served (permissive licence) · get_code("c39b7423940b40e6")
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 torch.utils.data as data

torch.manual_seed(0)

# Create sample data: list of tuples (image, caption)
# Images: (3, 256, 256) tensors
# Captions: variable length 1D tensors

sample_data = [
    (torch.randn(3, 256, 256), torch.tensor([1, 2, 3, 4, 5])),
    (torch.randn(3, 256, 256), torch.tensor([10, 11])),
]

ARGS = {
    "data": sample_data
}
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 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
define_models Ran
ran on a synthesized input
totalgood/viddesc/src/viddesc/keras_lstm.py
code served (permissive licence) · get_code("d3d0bc826c2e6782")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

ARGS = {
    'n_input': 4,
    'n_output': 3,
    'n_units': 8
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
keras, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
fib Ran
ran on a synthesized input
totalgood/viddesc/src/viddesc/skeleton.py
code served (permissive licence) · get_code("999ee8d595b9fdf7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned int, ndim 0, dtype family int, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function fib(n) takes an integer n and returns the n-th Fibonacci number
# No complex objects needed - just a simple integer parameter

ARGS = {
    "n": 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 int, ndim 0, dtype family int, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
float_feature Ran
ran on a synthesized input
totalgood/viddesc/src/viddesc/preprocess.py
code served (permissive licence) · get_code("44fda9771f751cfc")
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 Feature. 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 tensorflow as tf

# Analyze the function: float_feature(value)
# The function creates a tf.train.Feature with a float_list
# It expects 'value' to be an iterable of floats that can be passed to FloatList

# Create a small list of float values
value = [1.5, 2.3, 3.7, 4.1]

ARGS = {
    "value": value
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Feature — a Feature output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Feature 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
int64_feature Ran
ran; our contract draft was wrong, not the code
totalgood/viddesc/src/viddesc/preprocess.py
code served (permissive licence) · get_code("0c2d6e281d5fbeb5")
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 Feature. 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 tensorflow as tf

ARGS = {
    "value": 42
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Feature — a Feature output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Feature output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
load_embeddings Ran
ran on a synthesized input
totalgood/viddesc/src/viddesc/keras_lstm.py
code served (permissive licence) · get_code("5933f2701ea63955")
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 json
import os
import tempfile
import numpy as np
import shutil

# Create a temporary directory structure with mock data
temp_dir = tempfile.mkdtemp()

# Create a movie subdirectory
movie_dir = os.path.join(temp_dir, "movie_001")
os.makedirs(movie_dir)

# Create a sample JSON file with embeddings
sample_data = {
    "embeddings": [
        {
            "frame_id": 0,
            "mbox_loc": [10.5, 20.3, 100.2, 150.7]
        },
        {
            "frame_id": 1,
            "mbox_loc": [15.2, 25.1, 105.0, 155.5]
        }
    ]
}

json_file_path = os.path.join(movie_dir, "clip_001.json")
with open(json_file_path, 'w') as f:
    json.dump(sample_data, f)

# Create another movie subdirectory with another JSON file
movie_dir_2 = os.path.join(temp_dir, "movie_002")
os.makedirs(movie_dir_2)

sample_data_2 = {
    "embeddings": [
        {
            "frame_id": 2,
            "mbox_loc": [5.0, 10.0, 50.0, 100.0]
        }
    ]
}

json_file_path_2 = os.path.join(movie_dir_2, "clip_002.json")
with open(json_file_path_2, 'w') as f:
    json.dump(sample_data_2, f)

# Create a file with a dot in its name (should be skipped)
dotfile = os.path.join(temp_dir, ".hidden_file")
with open(dotfile, 'w') as f:
    f.write("should be skipped")

ARGS = {
    "path": temp_dir
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
bytes_feature Not yet run
unverified (harvested, no recorded run)
totalgood/viddesc/src/viddesc/preprocess.py
code served (permissive licence) · get_code("9e6c2c92eee0a29e")
How it ran

no invocation recorded.

resize_image Not yet run
unverified (harvested, no recorded run)
totalgood/viddesc/src/imgdesc/resize.py
code served (permissive licence) · get_code("ddccddb3e642d270")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 1 later) and 5 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
int64_feature 7 other papers
1 dated earlier, 6 later
1607.08022 (2016-07)
Instance Normalization: The Missing Ingredient for Fast Stylization
1802.05591 (2018-02)
Towards End-to-End Lane Detection: an Instance Segmentation Approach
1803.10562 (2018-03)
ELEGANT: Exchanging Latent Encodings with GAN for Transferring Multiple Face Attributes
and 4 more not named here
collate_fn 1 other paper
1 dated earlier
1411.4555 (2014-11)
Show and Tell: A Neural Image Caption Generator
resize_image 1 other paper
1 dated earlier
1411.4555 (2014-11)
Show and Tell: A Neural Image Caption Generator

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

While deep convolutional neural networks frequently approach or exceed human-level performance at benchmark tasks involving static images, extending this success to moving images is not straightforward. Having models which can learn to understand video is of interest for many applications, including content recommendation, prediction, summarization, event/object detection and understanding human visual perception, but many domains lack sufficient data to explore and perfect video models. In order to address the need for a simple, quantitative benchmark for developing and understanding video, we present MovieFIB, a fill-in-the-blank question-answering dataset with over 300,000 examples, based on descriptive video annotations for the visually impaired. In addition to presenting statistics and a description of the dataset, we perform a detailed analysis of 5 different models' predictions, and compare these with human performance. We investigate the relative importance of language, static (2D) visual features, and moving (3D) visual features; the effects of increasing dataset size, the number of frames sampled; and of vocabulary size. We illustrate that: this task is not solvable by a language model alone; our model combining 2D and 3D visual information indeed provides the best result; all models perform significantly worse than human-level. We provide human evaluations for responses given by different models and find that accuracy on the MovieFIB evaluation corresponds well with human judgement. We suggest avenues for improving video models, and hope that the proposed dataset can be useful for measuring and encouraging progress in this very interesting field.

For agents

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

get_harvested_code_for_paper("1611.07810")
get_code_for_paper("1611.07810")
have("1611.07810")

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

curl https://syntology.ai/api/ran/1611.07810.json

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

[![Syntology run record](https://syntology.ai/api/ran/1611.07810.svg)](https://syntology.ai/paper/1611.07810)

Connect an agent — have() is free.