SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.14418 · 2024

Uncertainty-Aware Evaluation for Vision-Language Models

arXiv · PDF · Open in the Atlas

Code that ran

8 of 17 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, 2 violated it and 0 have no contract verdict; 6 failures of our instrument; 9 no recorded run); 1 of them was also run on the host, not in the jail, by the peer-agreement lane that set its level, 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/2402.14418.json · badge: /api/ran/2402.14418.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
ensec-ai/vlm-uncertainty-bench canonical (classifier label)
classifier basis: plan_row
8 ran · 9 no recorded run
FunctionStatusWhere it lives
softmax Ran
ran, violated the contract we drafted
ensec-ai/vlm-uncertainty-bench/uncertainty_quantification_via_cp.py
code served (permissive licence) · get_code("7f9b15275ca5759c")
How it ran

Ran in Syntology's jail on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned ndarray, ndim 1, dtype family float, shape (4,) as a later run recorded it. The image it ran in was not recorded. The comparison behind this sample's level is a different run: the peer-agreement lane ran it again on the host, not in the jail (report dated 2026-09-05T00:37:48Z). 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

# The softmax function takes a single parameter 'x' which should be a numpy array
# It performs exponential normalization on the input

# Create a small test input - a 1D array
x = np.array([1.0, 2.0, 3.0, 4.0])

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 1, dtype family float, shape (4,) — shape (4,) 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_options Ran
ran; our contract draft was wrong, not the code
ensec-ai/vlm-uncertainty-bench/data_utils/common_utils.py
code served (permissive licence) · get_code("7ec8353674842469")
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 math

def is_none(value):
    if value is None:
        return True
    if type(value) is float and math.isnan(value):
        return True
    if type(value) is str and value.lower() == 'nan':
        return True
    return False

def get_options(row, options):
    parsed_options = []
    for option in options:
        option_value = row[option]
        if is_none(option_value):
            break
        parsed_options.append(option_value)
    return parsed_options

# Create a row dict with multiple option keys
row = {
    'option_1': 'value_1',
    'option_2': 'value_2',
    'option_3': 'value_3',
    'option_4': None,
}

# Create options list
options = ['option_1', 'option_2', 'option_3', 'option_4']

ARGS = {
    'row': row,
    'options': options,
}
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 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
is_none Ran
ran, violated the contract we drafted
ensec-ai/vlm-uncertainty-bench/data_utils/common_utils.py
code served (permissive licence) · get_code("25593cfba3f7b483")
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 math

def is_none(value):
    if value is None:
        return True
    if type(value) is float and math.isnan(value):
        return True
    if type(value) is str and value.lower() == 'nan':
        return True
    return False

# Test cases for the function
ARGS = {
    "value": None
}
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 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
open_json Ran
ran; our contract draft was wrong, not the code
ensec-ai/vlm-uncertainty-bench/data_utils/common_utils.py
code served (permissive licence) · get_code("8a064d582b5c57ea")
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 dict. 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 tempfile
import os

# Create a temporary JSON file for testing
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.json")

# Write a simple JSON file
test_data = {
    "key1": "value1",
    "key2": [1, 2, 3],
    "key3": {"nested": "object"}
}

with open(temp_file, 'w') as f:
    json.dump(test_data, f)

ARGS = {
    "path": temp_file
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
parse_img_path Ran
ran; our contract draft was wrong, not the code
ensec-ai/vlm-uncertainty-bench/data_utils/mmmu.py
code served (permissive licence) · get_code("de4427887bed9253")
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 re

def parse_img_path(text):
    matches = re.findall("<img='(.*?)'>", text)
    return matches

ARGS = {
    "text": "This is a test <img='path/to/image1.png'> and another <img='path/to/image2.jpg'> in the text."
}
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 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
process_ai2d Ran
ran; our contract draft was wrong, not the code
ensec-ai/vlm-uncertainty-bench/data_utils/ai2d.py
code served (permissive licence) · get_code("ca40ba6d177df427")
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 json
import glob
import tempfile
import shutil

# Create a temporary directory structure that mimics AI2D dataset
temp_dir = tempfile.mkdtemp()
questions_dir = os.path.join(temp_dir, 'questions')
images_dir = os.path.join(temp_dir, 'images')
os.makedirs(questions_dir)
os.makedirs(images_dir)

# Create sample question files
sample_questions = {
    'What is the color?': {
        'questionId': 'q1',
        'answerTexts': ['red', 'blue', 'green', 'yellow'],
        'correctAnswer': 0
    },
    'What is the shape?': {
        'questionId': 'q2',
        'answerTexts': ['circle', 'square', 'triangle', 'diamond'],
        'correctAnswer': 2
    }
}

# Write question files
for i in range(2):
    question_file = os.path.join(questions_dir, f'questions_{i}.json')
    data = {
        'imageName': f'image_{i}.png',
        'questions': sample_questions
    }
    with open(question_file, 'w') as f:
        json.dump(data, f)

# Create dummy image files
for i in range(2):
    image_path = os.path.join(images_dir, f'image_{i}.png')
    with open(image_path, 'w') as f:
        f.write('dummy image content')

ARGS = {
    'dataset_path': temp_dir
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tqdm — 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
process_seedbench Ran
ran; our contract draft was wrong, not the code
ensec-ai/vlm-uncertainty-bench/data_utils/seedbench.py
code served (permissive licence) · get_code("a1bac8b55fe3309a")
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 json
import tempfile
import shutil

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

# Create a mock SEED-Bench.json file with the expected structure
seed_bench_data = {
    "questions": [
        {
            "question_id": "q1",
            "question_type_id": 5,
            "data_id": "image_001.jpg",
            "choice_a": "Option A text",
            "choice_b": "Option B text",
            "choice_c": "Option C text",
            "choice_d": "Option D text"
        },
        {
            "question_id": "q2",
            "question_type_id": 8,
            "data_id": "image_002.jpg",
            "choice_a": "Another A",
            "choice_b": "Another B",
            "choice_c": "Another C",
            "choice_d": "Another D"
        },
        {
            "question_id": "q3",
            "question_type_id": 15,  # This will be filtered out (>= 10)
            "data_id": "image_003.jpg",
            "choice_a": "Filtered A",
            "choice_b": "Filtered B",
            "choice_c": "Filtered C",
            "choice_d": "Filtered D"
        }
    ]
}

# Create the image subdirectory
image_dir = os.path.join(temp_dir, 'SEED-Bench-image')
os.makedirs(image_dir, exist_ok=True)

# Write the SEED-Bench.json file
downloaded_file_path = os.path.join(temp_dir, 'SEED-Bench.json')
with open(downloaded_file_path, 'w') as f:
    json.dump(seed_bench_data, f)

# Create dummy image files
for i in range(1, 4):
    image_path = os.path.join(image_dir, f'image_00{i}.jpg')
    with open(image_path, 'w') as f:
        f.write('dummy image data')

ARGS = {
    'downloaded_file_path': downloaded_file_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 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
process_single_sample Ran
ran; our contract draft was wrong, not the code
ensec-ai/vlm-uncertainty-bench/data_utils/mmmu.py
code served (permissive licence) · get_code("c1b3610ee2583172")
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 dict. 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 re

def parse_img_path(text):
    matches = re.findall("<img='(.*?)'>", text)
    return matches

def process_single_sample(data):
    question = data['question']
    o_imgs_paths = []
    for option in data['options']:
        current_o_imgs_paths = parse_img_path(option)
        for img_path in current_o_imgs_paths:
            o_imgs_paths.append(img_path)
    if len(o_imgs_paths) > 1:
        return {'id': data['id'], 'question': question, 'options': data['options'], 'answer': data['answer'], 'image': None, 'question_type': data['question_type']}
    else:
        return {'id': data['id'], 'question': question, 'options': data['options'], 'answer': data['answer'], 'image': data['image_1'], 'img_type': data['img_type'], 'hint': None, 'topic_difficulty': data['topic_difficulty'], 'subfield': data['subfield']}

ARGS = {
    'data': {
        'id': 'sample_001',
        'question': 'What is the capital of France?',
        'options': [
            'Paris is the answer',
            'London is not',
            'Berlin is also not'
        ],
        'answer': 'Paris',
        'image_1': 'path/to/image.jpg',
        'img_type': 'jpg',
        'topic_difficulty': 'easy',
        'subfield': 'geography',
        'question_type': 'multiple_choice'
    }
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
download_ai2d Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/data_utils/ai2d.py
code served (permissive licence) · get_code("1267ef7cf3135661")
How it ran

no invocation recorded.

download_mmbench Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/data_utils/mmbench.py
code served (permissive licence) · get_code("a268ca9c33318e5c")
How it ran

no invocation recorded.

download_seedbench Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/data_utils/seedbench.py
code served (permissive licence) · get_code("25a29dad236f0992")
How it ran

no invocation recorded.

get_ce Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/uncertainty_quantification_via_cp.py
code served (permissive licence) · get_code("d40927dd38eca095")
How it ran

no invocation recorded.

get_dataset Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/data_utils/utils.py
code served (permissive licence) · get_code("9e203cb8767712bd")
How it ran

no invocation recorded.

get_logits Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/models_utils/utils.py
code served (permissive licence) · get_code("4335b7a5c5a94aaf")
How it ran

no invocation recorded.

process_mmmu Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/data_utils/mmmu.py
code served (permissive licence) · get_code("eba2dfc02d4349ea")
How it ran

no invocation recorded.

process_oodcv Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/data_utils/oodcv.py
code served (permissive licence) · get_code("5c69a8ac5c608ef9")
How it ran

no invocation recorded.

process_scienceqa Not yet run
unverified (harvested, no recorded run)
ensec-ai/vlm-uncertainty-bench/data_utils/scienceqa.py
code served (permissive licence) · get_code("d2a305808682d0b6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 17 samples lifted from this paper's linked repositories, 17 carry a function identity and 0 do not. Of the 17 distinct functions with an identity, 16 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 1 the same month, 16 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
softmax 12 other papers
5 dated earlier, 1 the same month, 6 later
1411.2738 (2014-11)
word2vec Parameter Learning Explained
1803.05337 (2018-03)
Learning to Recognize Musical Genre from Audio
1906.05394 (2019-06)
Neural Arabic Question Answering
and 9 more not named here
parse_img_path 5 other papers
5 later
2406.05343 (2024-06)
M3GIA: A Cognition Inspired Multilingual and Multimodal General Intelligence Ability Benchmark
Yang_PVC_Progressive_Visual_Token_Compression_for_Unified_Image_and_Video_CVPR_2025_paper (2025 (year only))
PVC: Progressive Visual Token Compression for Unified Image and Video Processing in Large Vision-Language Models
2502.00698 (2025-02)
MM-IQ: Benchmarking Human-Like Abstraction and Reasoning in Multimodal Models
and 2 more not named here
get_options 2 other papers
2 later
2411.16721 (2024-11)
Steering Away from Harm: An Adaptive Approach to Defending Vision Language Model Against Jailbreaks
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
download_ai2d 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
download_mmbench 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
download_seedbench 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
get_dataset 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
get_logits 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
is_none 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
open_json 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
process_ai2d 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
process_mmmu 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
process_oodcv 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
process_scienceqa 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
process_seedbench 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models
process_single_sample 1 other paper
1 later
2502.06884 (2025-02)
Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models

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

Vision-Language Models like GPT-4, LLaVA, and CogVLM have surged in popularity recently due to their impressive performance in several vision-language tasks. Current evaluation methods, however, overlook an essential component: uncertainty, which is crucial for a comprehensive assessment of VLMs. Addressing this oversight, we present a benchmark incorporating uncertainty quantification into evaluating VLMs. Our analysis spans 20+ VLMs, focusing on the multiple-choice Visual Question Answering (VQA) task. We examine models on 5 datasets that evaluate various vision-language capabilities. Using conformal prediction as an uncertainty estimation approach, we demonstrate that the models' uncertainty is not aligned with their accuracy. Specifically, we show that models with the highest accuracy may also have the highest uncertainty, which confirms the importance of measuring it for VLMs. Our empirical findings also reveal a correlation between model uncertainty and its language model part.

For agents

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

get_harvested_code_for_paper("2402.14418")
get_code_for_paper("2402.14418")
have("2402.14418")

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

curl https://syntology.ai/api/ran/2402.14418.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.14418.svg)](https://syntology.ai/paper/2402.14418)

Connect an agent — have() is free.