SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2408.07009 · 2024

Imagen 3

arXiv · PDF · Open in the Atlas

Code that ran

11 of 12 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; 3 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 8 failures of our instrument; 1 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/2408.07009.json · badge: /api/ran/2408.07009.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
linzhiqiu/t2i_metrics canonical (classifier label)
classifier basis: census
11 ran · 0 no recorded run
linzhiqiu/t2v_metrics canonical (classifier label)
classifier basis: plan_row
0 ran · 1 no recorded run
FunctionStatusWhere it lives
concatenate_images_grid Ran
ran; our contract draft was wrong, not the code
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/mm_utils.py
code served (permissive licence) · get_code("e00132d401b8d151")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family int, shape (600, 800, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import cv2
import numpy as np
from math import ceil, sqrt

# Create small test images with distinct dimensions
# Image 1: 100x80 (height x width)
img1 = np.random.randint(0, 256, (100, 80, 3), dtype=np.uint8)

# Image 2: 120x90 (height x width)
img2 = np.random.randint(0, 256, (120, 90, 3), dtype=np.uint8)

# Image 3: 110x85 (height x width)
img3 = np.random.randint(0, 256, (110, 85, 3), dtype=np.uint8)

# Image 4: 95x75 (height x width)
img4 = np.random.randint(0, 256, (95, 75, 3), dtype=np.uint8)

images = [img1, img2, img3, img4]
dist_images = 10
output_size = (800, 600)

ARGS = {
    "images": images,
    "dist_images": dist_images,
    "output_size": output_size
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family int, shape (600, 800, 3) — shape (600, 800, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
concatenate_images_horizontal Ran
ran, honoured the contract we drafted
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/mm_utils.py
code served (permissive licence) · get_code("3f916b77e8b386b3")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family int, shape (12, 63, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create small test images with distinct sizes
# Image 1: 10x15 (height x width)
img1 = np.random.randint(0, 256, (10, 15, 3), dtype=np.uint8)

# Image 2: 12x20 (different height and width)
img2 = np.random.randint(0, 256, (12, 20, 3), dtype=np.uint8)

# Image 3: 8x18 (yet another size)
img3 = np.random.randint(0, 256, (8, 18, 3), dtype=np.uint8)

images = [img1, img2, img3]

# Distance between images in pixels
dist_images = 5

ARGS = {
    "images": images,
    "dist_images": dist_images
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family int, shape (12, 63, 3) — shape (12, 63, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
concatenate_images_vertical Ran
ran, honoured the contract we drafted
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/mm_utils.py
code served (permissive licence) · get_code("07594707c816a98a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family int, shape (16, 6, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create small test images with distinct dimensions
# Image 1: 5 rows x 4 columns x 3 channels
img1 = np.random.randint(0, 256, size=(5, 4, 3), dtype=np.uint8)

# Image 2: 3 rows x 6 columns x 3 channels
img2 = np.random.randint(0, 256, size=(3, 6, 3), dtype=np.uint8)

# Image 3: 4 rows x 5 columns x 3 channels
img3 = np.random.randint(0, 256, size=(4, 5, 3), dtype=np.uint8)

images = [img1, img2, img3]
dist_images = 2

ARGS = {
    "images": images,
    "dist_images": dist_images
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family int, shape (16, 6, 3) — shape (16, 6, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
encode_image Ran
ran; our contract draft was wrong, not the code
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/gpt4v_model.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from jongwoopark7978/LVNet, is pointer only (licence: NONE) · get_code("f41cb1a19b154297")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from jongwoopark7978/LVNet, 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.

encode_image Ran
ran; our contract draft was wrong, not the code
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/gemini_model.py
code served (permissive licence) · get_code("2c2b1eb6c4d90f6c")
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 bytes. 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 tempfile
import os

# Create a temporary image file for testing
temp_dir = tempfile.mkdtemp()
image_path = os.path.join(temp_dir, "test_image.png")

# Write some binary data to simulate a real image file
with open(image_path, 'wb') as f:
    f.write(b'\x89PNG\r\n\x1a\n' + b'\x00' * 100)

ARGS = {
    'image_path': image_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bytes — a bytes output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a bytes 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
extract_frames Ran
ran; our contract draft was wrong, not the code
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/gemini_model.py
code served (permissive licence) · get_code("46c6c4b62cdd0bdb")
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, 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 cv2
import numpy as np
import tempfile
import os

# Create a temporary video file for testing
def create_test_video(video_path, num_frames_to_create=10, width=320, height=240):
    """Create a simple test video file"""
    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
    out = cv2.VideoWriter(video_path, fourcc, 20.0, (width, height))
    
    for i in range(num_frames_to_create):
        # Create a simple frame with changing color
        frame = np.zeros((height, width, 3), dtype=np.uint8)
        frame[:, :] = (i * 25 % 256, i * 50 % 256, i * 100 % 256)
        out.write(frame)
    
    out.release()

# Create temporary video file
temp_dir = tempfile.gettempdir()
video_path = os.path.join(temp_dir, "test_video_temp.mp4")
create_test_video(video_path, num_frames_to_create=20)

ARGS = {
    "video_path": video_path,
    "num_frames": 5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list 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
extract_frames Ran
ran; our contract draft was wrong, not the code
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/gpt4v_model.py
code served (permissive licence) · get_code("0037c0a0dea955f5")
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, 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 cv2
import numpy as np
import tempfile
import os

# Create a temporary video file for testing
temp_dir = tempfile.mkdtemp()
video_path = os.path.join(temp_dir, "test_video.mp4")

# Create a simple video file using opencv
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
out = cv2.VideoWriter(video_path, fourcc, 30.0, (640, 480))

# Write 30 frames to the video
for i in range(30):
    frame = np.zeros((480, 640, 3), dtype=np.uint8)
    # Add some variation to each frame
    frame[:, :] = [i * 8 % 255, (i * 4) % 255, (i * 2) % 255]
    out.write(frame)

out.release()

ARGS = {
    "video_path": video_path,
    "num_frames": 5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list 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
get_image_type Ran
ran; our contract draft was wrong, not the code
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/gemini_model.py
code served (permissive licence) · get_code("e1feb126f7f60749")
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 os
import tempfile

# Create a temporary directory and file to avoid actual file I/O
temp_dir = tempfile.gettempdir()

# Define valid image paths with different supported extensions
image_paths = [
    os.path.join(temp_dir, "test_image.png"),
    os.path.join(temp_dir, "test_image.jpg"),
    os.path.join(temp_dir, "test_image.jpeg"),
    os.path.join(temp_dir, "test_image.gif"),
    os.path.join(temp_dir, "test_image.bmp"),
    os.path.join(temp_dir, "test_image.webp"),
]

# Use the first valid image path
ARGS = {
    "image_path": image_paths[0]
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — 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_type Ran
ran; our contract draft was wrong, not the code
linzhiqiu/t2i_metrics/t2v_metrics/models/vqascore_models/gpt4v_model.py
code served (permissive licence) · get_code("e1f1f20eec0ebdf6")
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
import numpy as np

ARGS = {
    'image_path': 'test_image.png'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type 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
image_loader Ran
ran, honoured the contract we drafted
linzhiqiu/t2i_metrics/t2v_metrics/models/model.py
code served (permissive licence) · get_code("f13b1160fd3554a8")
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 Image. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import tempfile
import os

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

# Create a test .npy file with RGB image data (H x W x C format)
npy_path = os.path.join(temp_dir, 'test_image.npy')
npy_data = np.random.randint(0, 256, (4, 3, 3), dtype=np.uint8)  # 4x3 image with 3 channels
np.save(npy_path, npy_data)

# Create a test .jpg file using PIL
from PIL import Image
jpg_path = os.path.join(temp_dir, 'test_image.jpg')
test_img = Image.new('RGB', (4, 3), color=(73, 109, 137))
test_img.save(jpg_path)

# Test with npy file path
ARGS = {
    'image_path': npy_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
read_video_cv2 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
linzhiqiu/t2i_metrics/t2v_metrics/models/video_utils.py
code served (permissive licence) · get_code("9f8cc605dbfd5b5b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 4, dtype family int, shape (4, 240, 320, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import cv2
import numpy as np
from PIL import Image
import tempfile
import os

# Create a temporary video file for testing
def create_test_video(filepath, num_frames=10, width=320, height=240):
    """Create a simple test video with solid color frames"""
    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
    out = cv2.VideoWriter(filepath, fourcc, 20.0, (width, height))
    
    for i in range(num_frames):
        # Create a frame with a color that changes based on frame index
        frame = np.zeros((height, width, 3), dtype=np.uint8)
        frame[:, :, 0] = (i * 25) % 256  # Blue channel
        frame[:, :, 1] = (i * 50) % 256  # Green channel
        frame[:, :, 2] = (i * 75) % 256  # Red channel
        out.write(frame)
    
    out.release()

# Create temporary video file
temp_dir = tempfile.mkdtemp()
video_path = os.path.join(temp_dir, "test_video.mp4")
create_test_video(video_path, num_frames=10)

# Define arguments
ARGS = {
    "video_path": video_path,
    "all_indices": [0, 2, 4, 6]
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 4, dtype family int, shape (4, 240, 320, 3) — shape (4, 240, 320, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_conv_template Not yet run
unverified (harvested, no recorded run)
linzhiqiu/t2v_metrics/t2v_metrics/models/vqascore_models/fastchat_utils.py
code served (permissive licence) · get_code("4b4a1446c2b464c6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 3 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 2 the same month, 3 later) and 9 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
encode_image 76 other papers
26 dated earlier, 2 the same month, 47 later, 1 same year, order not determinable
2309.05950 (2023-09)
Language Models as Black-Box Optimizers for Vision-Language Models
2311.15732 (2023-11)
GPT4Vis: What Can GPT-4 Do for Zero-shot Visual Recognition?
2312.14867 (2023-12)
VIEScore: Towards Explainable Metrics for Conditional Image Synthesis Evaluation
and 73 more not named here
get_conv_template 69 other papers
38 dated earlier, 1 the same month, 25 later, 4 same year, order not determinable, 1 undated
1911.03705 (2019-11)
CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning
2112.05682 (2021-12)
Self-attention Does Not Need $O(n^2)$ Memory
2204.07705 (2022-04)
SUPER-NATURALINSTRUCTIONS: Generalization via Declarative Instructions on 1600+ NLP Tasks
and 66 more not named here
image_loader 1 other paper
1 later
2412.21059 (2024-12)
VisionReward: Fine-Grained Multi-Dimensional Human Preference Learning for Image and Video Generation

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We introduce Imagen 3, a latent diffusion model that generates high quality images from text prompts. We describe our quality and responsibility evaluations. Imagen 3 is preferred over other state-of-the-art (SOTA) models at the time of evaluation. In addition, we discuss issues around safety and representation, as well as methods we used to minimize the potential harm of our models.

For agents

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

get_harvested_code_for_paper("2408.07009")
get_code_for_paper("2408.07009")
have("2408.07009")

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

curl https://syntology.ai/api/ran/2408.07009.json

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

[![Syntology run record](https://syntology.ai/api/ran/2408.07009.svg)](https://syntology.ai/paper/2408.07009)

Connect an agent — have() is free.