SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1909.05863 · 2019

Finding Generalizable Evidence by Learning to Convince Q&A Models

arXiv · PDF · Open in the Atlas

Code that ran

11 of 22 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; 1 honoured the contract we drafted, 2 violated it and 2 have no contract verdict; 6 failures of our instrument; 11 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/1909.05863.json · badge: /api/ran/1909.05863.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
ethanjperez/convince canonical (classifier label)
classifier basis: census
11 ran · 11 no recorded run
FunctionStatusWhere it lives
built Ran
ran, violated the contract we drafted
ethanjperez/convince/ParlAI/parlai/core/build_data.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from SivilTaram/Persona-Dialogue-Generation, is pointer only (licence: MIT) · get_code("5171451a0030de60")
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 SivilTaram/Persona-Dialogue-Generation, and MIT, as GitHub detects it on that repository, is in the permissive set, but a licence decision keeps the body a pointer: its licence file names an outside party, so it may cover vendored code rather than the repository owner's files (recorded as pointer:licence_names_third_party_corporation). 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.

default_collate Ran
ran; our contract draft was wrong, not the code
ethanjperez/convince/ParlAI/parlai/core/pytorch_data_teacher.py
code served (permissive licence) · get_code("a9ef7e3192d9397e")
How it ran

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

import torch

torch.manual_seed(0)

# The function takes a batch parameter which is iterated over.
# Each element b in batch is expected to be a tuple where:
#   - b[0] is an index (idx)
#   - b[1] is either a list (and we take b[1][0]) or something else (ep = b[1])
# The function constructs new_batch as a list of (idx, ep) tuples

# Create sample batch data
batch_data = [
    (0, [torch.tensor([1.0, 2.0, 3.0, 4.0])]),  # idx=0, b[1] is a list
    (1, torch.tensor([5.0, 6.0, 7.0, 8.0])),     # idx=1, b[1] is a tensor directly
    (2, [torch.tensor([9.0, 10.0, 11.0, 12.0])]),  # idx=2, b[1] is a list
]

ARGS = {
    "batch": batch_data
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
escape Ran
ran; our contract draft was wrong, not the code
ethanjperez/convince/ParlAI/parlai/core/dict.py
code served (permissive licence) · get_code("eeaa9721dd6b0c0e")
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

ARGS = {
    's': 'hello\nworld\ttab\rcarriage'
}
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
find_ngrams Ran
ran; our contract draft was wrong, not the code
ethanjperez/convince/ParlAI/parlai/core/dict.py
code served (permissive licence) · get_code("63c2a8c36bd5cbff")
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.

# Setup code for find_ngrams function

# Construct arguments for find_ngrams
token_dict = {
    'hello world': True,
    'world of': True,
    'of python': True,
    'python is': True,
    'is great': True,
    'hello': True,
    'world': True,
    'of': True,
    'python': True,
    'is': True,
    'great': True,
}

text = ['hello', 'world', 'of', 'python', 'is', 'great']

n = 3

ARGS = {
    'token_dict': token_dict,
    'text': text,
    'n': n,
}
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_tensor Ran
ran on a synthesized input
ethanjperez/convince/ParlAI/parlai/core/thread_utils.py
code served (permissive licence) · get_code("9fec83adf9b3df46")
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
import torch.nn as nn

torch.manual_seed(0)

# The function is_tensor takes a single parameter 'v' and checks if it's a torch tensor
# by examining its type's module and calling torch.is_tensor on it

# Test with an actual torch tensor
tensor_arg = torch.randn(2, 4)

# Create the ARGS dict
ARGS = {
    'v': tensor_arg
}
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
modelzoo_path Ran
ran on a synthesized input
ethanjperez/convince/ParlAI/parlai/core/build_data.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from SivilTaram/Persona-Dialogue-Generation, is pointer only (licence: MIT) · get_code("3aee60807f1b6758")
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 SivilTaram/Persona-Dialogue-Generation, and MIT, as GitHub detects it on that repository, is in the permissive set, but a licence decision keeps the body a pointer: its licence file names an outside party, so it may cover vendored code rather than the repository owner's files (recorded as pointer:licence_names_third_party_corporation). 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.

name_to_agent_class Ran
ran; our contract draft was wrong, not the code
ethanjperez/convince/ParlAI/parlai/core/agents.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from SivilTaram/Persona-Dialogue-Generation, is pointer only (licence: MIT) · get_code("e3188064e3152f36")
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 SivilTaram/Persona-Dialogue-Generation, and MIT, as GitHub detects it on that repository, is in the permissive set, but a licence decision keeps the body a pointer: its licence file names an outside party, so it may cover vendored code rather than the repository owner's files (recorded as pointer:licence_names_third_party_corporation). 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.

normalize_answer Ran
ran; our contract draft was wrong, not the code
ethanjperez/convince/ParlAI/parlai/core/metrics.py
code served (permissive licence) · get_code("38a6695b805ebd76")
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 re

re_art = re.compile('\\b(a|an|the)\\b')
re_punc = re.compile('[!"#$%&()*+,-./:;<=>?@\\[\\]\\\\^`{|}~_\\\']')

def normalize_answer(s):
    """Lower text and remove punctuation, articles and extra whitespace."""

    def remove_articles(text):
        return re_art.sub(' ', text)

    def white_space_fix(text):
        return ' '.join(text.split())

    def remove_punc(text):
        return re_punc.sub(' ', text)

    def lower(text):
        return text.lower()
    return white_space_fix(remove_articles(remove_punc(lower(s))))


# The function takes a single parameter: s (a string)
# It processes the string through lower(), remove_punc(), remove_articles(), and white_space_fix()
# All operations are string operations

ARGS = {
    's': 'The Quick Brown Fox! A lazy dog, and the cat.'
}
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
str2bool Ran
ran, violated the contract we drafted
ethanjperez/convince/ParlAI/parlai/core/params.py
code served (permissive licence) · get_code("7ebbd59995cd9606")
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 argparse

def str2bool(value):
    v = value.lower()
    if v in ('yes', 'true', 't', '1', 'y'):
        return True
    elif v in ('no', 'false', 'f', 'n', '0'):
        return False
    else:
        raise argparse.ArgumentTypeError('Boolean value expected.')

ARGS = {
    'value': 'true'
}
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
str2floats Ran
ran, honoured the contract we drafted
ethanjperez/convince/ParlAI/parlai/core/params.py
code served (permissive licence) · get_code("69dc699488f428ae")
How it ran

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

import torch

torch.manual_seed(0)

ARGS = {
    's': '1.5,2.3,3.7'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
unescape Ran
ran; our contract draft was wrong, not the code
ethanjperez/convince/ParlAI/parlai/core/dict.py
code served (permissive licence) · get_code("f38682c3d9895af5")
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.

ARGS = {
    's': 'Hello\\nWorld\\tTab\\rReturn'
}
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
all_gather_list Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/distributed_utils.py
code served (permissive licence) · get_code("e50b83caf354e9ba")
How it ran

no invocation recorded.

ep_length Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/pytorch_data_teacher.py
code served (permissive licence) · get_code("65dd92b1f8b84c90")
How it ran

no invocation recorded.

first_n_cache Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/image_featurizers.py
code served (permissive licence) · get_code("6bba3a220a077794")
How it ran

no invocation recorded.

get_agent_module Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/agents.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from SivilTaram/Persona-Dialogue-Generation, is pointer only (licence: MIT) · get_code("617b91a1760c2f60")
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 SivilTaram/Persona-Dialogue-Generation, and MIT, as GitHub detects it on that repository, is in the permissive set, but a licence decision keeps the body a pointer: its licence file names an outside party, so it may cover vendored code rather than the repository owner's files (recorded as pointer:licence_names_third_party_corporation). 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. No invocation recorded.

get_model_name Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/params.py
code served (permissive licence) · get_code("7e67cccea94d1090")
How it ran

no invocation recorded.

load_agent_module Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/agents.py
code served (permissive licence) · get_code("e30ba42f67f26be1")
How it ran

no invocation recorded.

override_print Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/distributed_utils.py
code served (permissive licence) · get_code("2fd490d31e87fd01")
How it ran

no invocation recorded.

skipIfGPU Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/testing_utils.py
code served (permissive licence) · get_code("a1cb2ebdf97dd31f")
How it ran

no invocation recorded.

skipUnlessGPU Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/testing_utils.py
code served (permissive licence) · get_code("60108732130f0948")
How it ran

no invocation recorded.

skipUnlessTorch Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/testing_utils.py
code served (permissive licence) · get_code("5ad8c9ee18878de9")
How it ran

no invocation recorded.

sync_object Not yet run
unverified (harvested, no recorded run)
ethanjperez/convince/ParlAI/parlai/core/distributed_utils.py
code served (permissive licence) · get_code("9c66f5aea1de9db1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 22 samples lifted from this paper's linked repositories, 22 carry a function identity and 0 do not. Of the 22 distinct functions with an identity, 13 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 13 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
built 3 other papers
1 dated earlier, 2 later
1906.01603 (2019-06)
Do Neural Dialog Systems Use the Conversation History Effectively? An Empirical Study
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
2004.05388 (2020-04)
You Impress Me: Dialogue Generation via Mutual Persona Perception
escape 3 other papers
1 dated earlier, 2 later
1906.01603 (2019-06)
Do Neural Dialog Systems Use the Conversation History Effectively? An Empirical Study
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
2212.07919 (2022-12)
ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoning
find_ngrams 3 other papers
1 dated earlier, 2 later
1906.01603 (2019-06)
Do Neural Dialog Systems Use the Conversation History Effectively? An Empirical Study
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
2212.07919 (2022-12)
ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoning
unescape 3 other papers
1 dated earlier, 2 later
1906.01603 (2019-06)
Do Neural Dialog Systems Use the Conversation History Effectively? An Empirical Study
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
2212.07919 (2022-12)
ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoning
get_agent_module 2 other papers
2 later
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
2004.05388 (2020-04)
You Impress Me: Dialogue Generation via Mutual Persona Perception
name_to_agent_class 2 other papers
2 later
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
2004.05388 (2020-04)
You Impress Me: Dialogue Generation via Mutual Persona Perception
str2bool 2 other papers
1 dated earlier, 1 later
1906.01603 (2019-06)
Do Neural Dialog Systems Use the Conversation History Effectively? An Empirical Study
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
default_collate 1 other paper
1 later
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
ep_length 1 other paper
1 later
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
load_agent_module 1 other paper
1 later
2004.05388 (2020-04)
You Impress Me: Dialogue Generation via Mutual Persona Perception
modelzoo_path 1 other paper
1 later
2004.05388 (2020-04)
You Impress Me: Dialogue Generation via Mutual Persona Perception
normalize_answer 1 other paper
1 later
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generation
str2floats 1 other paper
1 later
2003.11963 (2020-03)
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance 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 propose a system that finds the strongest supporting evidence for a given answer to a question, using passage-based question-answering (QA) as a testbed. We train evidence agents to select the passage sentences that most convince a pretrained QA model of a given answer, if the QA model received those sentences instead of the full passage. Rather than finding evidence that convinces one model alone, we find that agents select evidence that generalizes; agent-chosen evidence increases the plausibility of the supported answer, as judged by other QA models and humans. Given its general nature, this approach improves QA in a robust manner: using agent-selected evidence (i) humans can correctly answer questions with only ~20% of the full passage and (ii) QA models can generalize to longer passages and harder questions.

For agents

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

get_harvested_code_for_paper("1909.05863")
get_code_for_paper("1909.05863")
have("1909.05863")

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

curl https://syntology.ai/api/ran/1909.05863.json

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

[![Syntology run record](https://syntology.ai/api/ran/1909.05863.svg)](https://syntology.ai/paper/1909.05863)

Connect an agent — have() is free.