SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2111.09734 · 2021

ClipCap: CLIP Prefix for Image Captioning

arXiv · PDF · Open in the Atlas

Code that ran

6 of 6 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 3 have no contract verdict; 3 failures of our instrument; 0 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/2111.09734.json · badge: /api/ran/2111.09734.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
rmokady/clip_prefix_caption canonical (classifier label)
classifier basis not recorded on the link
5 ran · 0 no recorded run
sithu31296/image-captioning reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
FunctionStatusWhere it lives
add_period Ran
ran on a synthesized input
rmokady/clip_prefix_caption/parse_conceptual.py
code served (permissive licence) · get_code("c68cb2401825d6a1")
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 str. 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

def add_period(caption: str):
    caption = caption.strip()
    if caption[-1] != '.':
        caption = caption + '.'
    elif caption[-2] == ' ':
        caption = caption[:-2] + '.'
    return caption

ARGS = {
    'caption': 'hello world'
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
generate2 Ran
ran; our contract draft was wrong, not the code
rmokady/clip_prefix_caption/predict.py
pointer only (licence: MIT); the stored copy get_code serves, from chuangchuangtan/c2p-clip-deepfakedetection, is pointer only (licence: NONE) · get_code("e33b112514a9e4b0")
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 chuangchuangtan/c2p-clip-deepfakedetection, 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.

generate_beam Ran
ran; our contract draft was wrong, not the code
rmokady/clip_prefix_caption/predict.py
pointer only (licence: MIT); the stored copy get_code serves, from yvsriram/FACTS, is pointer only (licence: NONE) · get_code("883320eb62f15d12")
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 yvsriram/FACTS, 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.

get_annotation_file Ran
ran; our contract draft was wrong, not the code
sithu31296/image-captioning/tools/val.py
code served (permissive licence) · get_code("6800239cecf973e8")
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 str. 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

ARGS = {
    'dataset': 'COCO'
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_image Ran
ran on a synthesized input
rmokady/clip_prefix_caption/parse_conceptual.py
code served (permissive licence) · get_code("0a286fb23ba005b9")
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 bool, 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 requests
import shutil
import tempfile
import os
from unittest.mock import Mock, patch

# Create a temporary directory for output
temp_dir = tempfile.mkdtemp()
out_path = os.path.join(temp_dir, "test_image.jpg")

# Mock the requests.get to avoid actual network calls
# We'll set up valid arguments that would work if the network call succeeded
url = "https://example.com/image.jpg"
timeout = 10

ARGS = {
    "url": url,
    "out_path": out_path,
    "timeout": timeout
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
requests — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, 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
thread Ran
ran on a synthesized input
rmokady/clip_prefix_caption/parse_conceptual.py
code served (permissive licence) · get_code("50449115c68a0ddb")
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 tempfile
import threading
from tqdm import tqdm
import os

# Create a temporary directory for output
temp_dir = tempfile.mkdtemp()

# Create the subdirectory for images
suffix = "images"
out_root = os.path.join(temp_dir, suffix)
os.makedirs(out_root, exist_ok=True)

# Construct the urls parameter: List[Tuple[List[str], int]]
# Each element is a tuple of (caption_url_pair, index)
# where caption_url_pair is [caption, url]
urls = [
    (["cat photo", "http://example.com/cat1.jpg"], 0),
    (["dog photo", "http://example.com/dog1.jpg"], 1),
    (["bird photo", "http://example.com/bird1.jpg"], 2),
]

# Thread ID
thread_id = 0

# Progress bar
progress = tqdm(total=len(urls), desc="Processing")

# Lock for thread safety
lock = threading.Lock()

# Suffix for output files
suffix = "images"

# Conceptual root directory
conceptual_root = temp_dir

ARGS = {
    "urls": urls,
    "thread_id": thread_id,
    "progress": progress,
    "lock": lock,
    "suffix": suffix,
    "conceptual_root": conceptual_root,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
requests, tqdm — 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

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 5 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 5 later) and 1 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
generate2 4 other papers
4 later
2211.11694 (2022-11)
Exploring Discrete Diffusion Models for Image Captioning
2309.17430 (2023-09)
FACTS: First Amplify Correlations and Then Slice to Discover Bias
2408.09647 (2024-08)
C2P-CLIP: Injecting Category Common Prompt in CLIP to Enhance Generalization in Deepfake Detection
and 1 more not named here
generate_beam 3 other papers
3 later
2211.11694 (2022-11)
Exploring Discrete Diffusion Models for Image Captioning
2309.17430 (2023-09)
FACTS: First Amplify Correlations and Then Slice to Discover Bias
2412.20206 (2024-12)
Toward Visual Grounding: A Survey
add_period 2 other papers
2 later
2211.11694 (2022-11)
Exploring Discrete Diffusion Models for Image Captioning
2412.20206 (2024-12)
Toward Visual Grounding: A Survey
get_image 2 other papers
2 later
2211.11694 (2022-11)
Exploring Discrete Diffusion Models for Image Captioning
2412.20206 (2024-12)
Toward Visual Grounding: A Survey
thread 2 other papers
2 later
2211.11694 (2022-11)
Exploring Discrete Diffusion Models for Image Captioning
2412.20206 (2024-12)
Toward Visual Grounding: A Survey

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

Image captioning is a fundamental task in vision-language understanding, where the model predicts a textual informative caption to a given input image. In this paper, we present a simple approach to address this task. We use CLIP encoding as a prefix to the caption, by employing a simple mapping network, and then fine-tunes a language model to generate the image captions. The recently proposed CLIP model contains rich semantic features which were trained with textual context, making it best for vision-language perception. Our key idea is that together with a pre-trained language model (GPT2), we obtain a wide understanding of both visual and textual data. Hence, our approach only requires rather quick training to produce a competent captioning model. Without additional annotations or pre-training, it efficiently generates meaningful captions for large-scale and diverse datasets. Surprisingly, our method works well even when only the mapping network is trained, while both CLIP and the language model remain frozen, allowing a lighter architecture with less trainable parameters. Through quantitative evaluation, we demonstrate our model achieves comparable results to state-of-the-art methods on the challenging Conceptual Captions and nocaps datasets, while it is simpler, faster, and lighter. Our code is available in https://github.com/rmokady/CLIP_prefix_caption.

For agents

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

get_harvested_code_for_paper("2111.09734")
get_code_for_paper("2111.09734")
have("2111.09734")

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

curl https://syntology.ai/api/ran/2111.09734.json

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

[![Syntology run record](https://syntology.ai/api/ran/2111.09734.svg)](https://syntology.ai/paper/2111.09734)

Connect an agent — have() is free.