SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2407.11277 · 2024

Target conversation extraction: Source separation using turn-taking dynamics

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; 2 honoured the contract we drafted, 2 violated it and 0 have no contract verdict; 7 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/2407.11277.json · badge: /api/ran/2407.11277.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
chentuochao/target-conversation-extraction canonical (classifier label)
classifier basis: census
11 ran · 1 no recorded run
FunctionStatusWhere it lives
calculate_cosine_similarity Ran
ran, violated the contract we drafted
chentuochao/target-conversation-extraction/eval_conversation.py
code served (permissive licence) · get_code("a0208c2aa866774d")
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 float64, ndim 0, 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
from sklearn.metrics.pairwise import cosine_similarity

# Create two numpy vectors with the same shape
vector1 = np.array([1.0, 2.0, 3.0, 4.0])
vector2 = np.array([2.0, 3.0, 4.0, 5.0])

ARGS = {
    'vector1': vector1,
    'vector2': vector2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
filter_set_by_keys Ran
ran; our contract draft was wrong, not the code
chentuochao/target-conversation-extraction/datasets/dataset_loader.py
code served (permissive licence) · get_code("d427b8df3a56508a")
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 torch

torch.manual_seed(0)

# Construct arguments for filter_set_by_keys
original_set = {
    "key_a": torch.randn(2, 4),
    "key_b": torch.randn(3, 5),
    "key_c": 42,
    "key_d": "hello",
    "key_e": [1, 2, 3],
}

key_sublist = ["key_a", "key_c", "key_z"]  # key_z not in original_set

ARGS = {
    "original_set": original_set,
    "key_sublist": key_sublist,
}
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 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
get_envelope Ran
ran, violated the contract we drafted
chentuochao/target-conversation-extraction/convert_AMI2Libri.py
code served (permissive licence) · get_code("6083727b41952440")
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 1, dtype family float, shape (100,) 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 a simple 1D signal for testing
signal = np.random.randn(100)

ARGS = {
    "signal": signal
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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 (100,) — shape (100,) 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_random_mic_positions Ran
ran, honoured the contract we drafted
chentuochao/target-conversation-extraction/datasets/dataset_util.py
code served (permissive licence) · get_code("60539ba4504bfe7d")
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 2, dtype family float, shape (1, 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

np.random.seed(0)

ARGS = {
    "n_mics": 5,
    "left": 0.0,
    "right": 10.0,
    "bottom": 0.0,
    "top": 8.0,
}
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 2, dtype family float, shape (1, 3) — shape (1, 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
high_pass_filter Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
chentuochao/target-conversation-extraction/convert_AMI2Libri.py
code served (permissive licence) · get_code("e8c863bc61c7255a")
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 1, dtype family float, shape (100,) 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
from scipy.signal import butter, lfilter

# Create a small audio waveform
wav = np.random.randn(100).astype(np.float32)

# Define sampling rate
sampling_rate = 16000

# Define order parameter
order = 7

# Construct ARGS dict with all parameters
ARGS = {
    'wav': wav,
    'sampling_rate': sampling_rate,
    'order': order
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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 (100,) — shape (100,) 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
import_attr Ran
ran; our contract draft was wrong, not the code
chentuochao/target-conversation-extraction/src/utils.py
code served (permissive licence) · get_code("63c5cc839a76504a")
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 function. 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 importlib
import torch
import torch.nn as nn

# The function import_attr takes an import_path string and dynamically imports
# and returns an attribute from a module. It uses:
# - rsplit('.', 1) on the import_path string
# - importlib.import_module() to import the module
# - getattr() to get the attribute
# 
# We need to provide a valid import_path that points to an actual importable
# attribute from the Python standard library or torch.

# A simple valid example: import torch.nn.functional.relu
# This would be the string 'torch.nn.functional.relu'

ARGS = {
    'import_path': 'torch.nn.functional.relu'
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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_net Ran
ran; our contract draft was wrong, not the code
chentuochao/target-conversation-extraction/src/utils.py
code served (permissive licence) · get_code("f22b1d22bc27279d")
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 MinimalPLModule. 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 directory for our test files
temp_dir = tempfile.mkdtemp()

# Create a minimal PyTorch module to use as pl_module
import torch
import torch.nn as nn

class MinimalPLModule(nn.Module):
    def __init__(self, hidden_dim=4, **kwargs):
        super().__init__()
        self.linear = nn.Linear(hidden_dim, hidden_dim)
        self.init_ckpt = kwargs.get('init_ckpt', None)
    
    def forward(self, x):
        return self.linear(x)

# Register the module in a way that can be imported
import sys
class FakeModule:
    MinimalPLModule = MinimalPLModule

sys.modules['fake_pl_module'] = FakeModule()

# Create a config JSON file
config_data = {
    "pl_module": "fake_pl_module.MinimalPLModule",
    "pl_module_args": {
        "hidden_dim": 4,
        "init_ckpt": "some_checkpoint.pt"
    },
    "learning_rate": 0.001,
    "batch_size": 2
}

config_path = os.path.join(temp_dir, "config.json")
with open(config_path, 'w') as f:
    json.dump(config_data, f)

# Define the ARGS dictionary
ARGS = {
    "expriment_config": config_path,
    "return_params": False
}
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 MinimalPLModule — a MinimalPLModule output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MinimalPLModule 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
mixing Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
chentuochao/target-conversation-extraction/datasets/generate_dataset.py
code served (permissive licence) · get_code("a2ea924d9b7f0fa1")
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 numpy as np

def augment_reverbration(voice_lists, args, total_samples):
    """Stand-in function for reverberation augmentation"""
    # Return modified voice_lists and room/speaker info
    processed_voice_lists = [v for v in voice_lists]
    room_info = {"room_dim": [4, 5, 3]}
    spk_info = {"num_speakers": len(voice_lists)}
    return processed_voice_lists, room_info, spk_info

class ArgsConfig:
    """Minimal args config with required attributes"""
    def __init__(self):
        self.reverb = 0.3  # probability of reverberation
        self.total_samples = 16000

# Create sample audio arrays with non-zero content
sample_len = 16000

# Create dialogue structure
dialogue = {
    'self_speech': {
        'speech': np.random.randn(sample_len) * 0.5 + 0.1,  # Non-zero audio
        'clean_example': np.random.randn(sample_len) * 0.5 + 0.1
    },
    'others': [
        {
            'speech': np.random.randn(sample_len) * 0.5 + 0.1
        },
        {
            'speech': np.random.randn(sample_len) * 0.5 + 0.1
        }
    ]
}

# Create interference list
interference = [
    {'speech': np.random.randn(1, sample_len) * 0.3 + 0.05},
    {'speech': np.random.randn(1, sample_len) * 0.3 + 0.05}
]

# Create noise array
noise = np.random.randn(sample_len) * 0.1 + 0.01

# Create args
args = ArgsConfig()

# Define all arguments
ARGS = {
    'dialogue': dialogue,
    'intereference': interference,
    'noise': noise,
    'args': args,
    'target_name': 'target_speaker',
    'interfer_name': 'interfering_speaker'
}
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 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
moving_average Ran
ran; our contract draft was wrong, not the code
chentuochao/target-conversation-extraction/convert_AMI2Libri.py
code served (permissive licence) · get_code("032038059574b300")
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 1, dtype family float, shape (8,) 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

# Construct arguments for moving_average function
data = np.array([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0])
window_size = 3

ARGS = {
    "data": data,
    "window_size": window_size,
}
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 1, dtype family float, shape (8,) — shape (8,) 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
overlap_duration_check Ran
ran, honoured the contract we drafted
chentuochao/target-conversation-extraction/eval_conversation.py
code served (permissive licence) · get_code("6af82b66048558cc")
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 float, 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 numpy as np

def overlap_duration_check(self_speaker, others, shift=1):
    L_total = 16000 * 60
    self_time = self_speaker['timestamp']
    if shift != 0:
        for i in range(0, len(self_time)):
            self_time[i] = [self_time[i][0] + shift, self_time[i][1] + shift]
    timestamps = [self_time]
    search_index = [0]
    for o in others:
        timestamps.append(o['timestamp'])
        search_index.append(0)
    step = 4000
    overlap_duration = 0
    valid_duration = 0
    diag_info = {}
    duration = []
    for stamp in timestamps:
        du = 0
        for b, e in stamp:
            du += e - b
        duration.append(du)
    duration = np.array(duration)
    duration = duration / np.sum(duration)
    diag_info['occupation'] = duration.tolist()
    for i in range(0, L_total, step):
        begin = i
        end = i + step
        spk_exits = []
        for j in range(len(timestamps)):
            while True:
                idx = search_index[j]
                if idx >= len(timestamps[j]):
                    break
                b, e = timestamps[j][idx]
                if end <= b:
                    break
                elif e <= begin:
                    search_index[j] += 1
                    continue
                else:
                    spk_exits.append(j)
                    if e < end:
                        search_index[j] += 1
                    break
        if len(spk_exits) > 1:
            overlap_duration += step
        if len(spk_exits) > 0:
            valid_duration += step
    overlap_duration = overlap_duration / (valid_duration + 1e-05)
    return overlap_duration

# Create a minimal stand-in class to hold self
class DummySelf:
    pass

# Create arguments
self_speaker = {
    'timestamp': [[1000, 5000], [10000, 15000], [20000, 25000]]
}

others = [
    {'timestamp': [[2000, 6000], [12000, 18000]]},
    {'timestamp': [[3000, 8000], [14000, 22000]]}
]

shift = 1

ARGS = {
    'self_speaker': self_speaker,
    'others': others,
    'shift': shift
}
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 float, 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
point_in_box Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
chentuochao/target-conversation-extraction/eval_conversation.py
code served (permissive licence) · get_code("7e2c6b8c214572ae")
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 torch

torch.manual_seed(0)

# Analyze the function:
# - pos: a sequence/list with at least 2 elements (pos[0], pos[1])
# - left, right, top, bottom: numeric values for boundary comparison

# Create a simple position as a list or tuple with 2 coordinates
pos = [5, 10]

# Define boundary values
left = 0
right = 15
top = 20
bottom = 5

ARGS = {
    "pos": pos,
    "left": left,
    "right": right,
    "top": top,
    "bottom": bottom
}
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 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
load_net_torch Not yet run
unverified (harvested, no recorded run)
chentuochao/target-conversation-extraction/src/utils.py
code served (permissive licence) · get_code("8470ce8d00caec6c")
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, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 0 later) and 11 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
import_attr 1 other paper
1 dated earlier
2405.06289 (2024-05)
Look Once to Hear: Target Speech Hearing with Noisy Examples

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

Extracting the speech of participants in a conversation amidst interfering speakers and noise presents a challenging problem. In this paper, we introduce the novel task of target conversation extraction, where the goal is to extract the audio of a target conversation based on the speaker embedding of one of its participants. To accomplish this, we propose leveraging temporal patterns inherent in human conversations, particularly turn-taking dynamics, which uniquely characterize speakers engaged in conversation and distinguish them from interfering speakers and noise. Using neural networks, we show the feasibility of our approach on English and Mandarin conversation datasets. In the presence of interfering speakers, our results show an 8.19 dB improvement in signal-to-noise ratio for 2-speaker conversations and a 7.92 dB improvement for 2-4-speaker conversations. Code, dataset available at https://github.com/chentuochao/Target-Conversation-Extraction.

For agents

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

get_harvested_code_for_paper("2407.11277")
get_code_for_paper("2407.11277")
have("2407.11277")

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

curl https://syntology.ai/api/ran/2407.11277.json

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

[![Syntology run record](https://syntology.ai/api/ran/2407.11277.svg)](https://syntology.ai/paper/2407.11277)

Connect an agent — have() is free.