SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2003.10580 · 2020

Meta Pseudo Labels

arXiv · PDF · Open in the Atlas

Code that ran

12 of 14 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 7 failures of our instrument; 2 no recorded run), measured 2026-11-21; 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/2003.10580.json · badge: /api/ran/2003.10580.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
ifsheldon/MPL_Lightning community (classifier label)
classifier basis not recorded on the link
7 ran · 0 no recorded run
usccolumbia/tsdnn community (classifier label)
classifier basis not recorded on the link
5 ran · 2 no recorded run
FunctionStatusWhere it lives
AutoContrast Ran
ran; our contract draft was wrong, not the code
ifsheldon/MPL_Lightning/mpl_lightning/augmentation.py
pointer only (licence: MIT); the stored copy get_code serves, from MaxH1996/FedXDS, is pointer only (licence: NONE) · get_code("c6e7c3451c85404f")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from MaxH1996/FedXDS, 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.

Brightness Ran
ran; our contract draft was wrong, not the code
ifsheldon/MPL_Lightning/mpl_lightning/augmentation.py
pointer only (licence: MIT); the stored copy get_code serves, from MaxH1996/FedXDS, is pointer only (licence: NONE) · get_code("0d1a8d6ce2015fc7")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from MaxH1996/FedXDS, 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.

Color Ran
ran; our contract draft was wrong, not the code
ifsheldon/MPL_Lightning/mpl_lightning/augmentation.py
pointer only (licence: MIT); the stored copy get_code serves, from MaxH1996/FedXDS, is pointer only (licence: NONE) · get_code("32cb3f5760fa593e")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from MaxH1996/FedXDS, 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.

add_data_specific_args Ran
ran on a synthesized input
ifsheldon/MPL_Lightning/mpl_lightning/datamodules.py
code served (permissive licence) · get_code("cc7636585a82c554")
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 ArgumentParser. 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.

from argparse import ArgumentParser

torch_imported = True
try:
    import torch
    torch.manual_seed(0)
except ImportError:
    torch_imported = False

# Create a parent parser
parent_parser = ArgumentParser()

# ARGS dict mapping the parameter 'parent_parser' to the ArgumentParser instance
ARGS = {
    'parent_parser': parent_parser
}
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 ArgumentParser — a ArgumentParser output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ArgumentParser 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
build_wideresnet Ran
ran on a synthesized input
ifsheldon/MPL_Lightning/mpl_lightning/aux_modules.py
code served (permissive licence) · get_code("4aed6586a15047dd")
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 WideResNet. 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 logging

torch.manual_seed(0)
logging.basicConfig(level=logging.INFO)

class Args:
    def __init__(self, dataset, num_classes, dense_dropout, local_rank):
        self.dataset = dataset
        self.num_classes = num_classes
        self.dense_dropout = dense_dropout
        self.local_rank = local_rank

ARGS = {
    "args": Args(
        dataset="cifar10",
        num_classes=10,
        dense_dropout=0.5,
        local_rank=0
    )
}
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 WideResNet — a WideResNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a WideResNet 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
class_eval Ran
ran; our contract draft was wrong, not the code
usccolumbia/tsdnn/predict.py
code served (permissive licence) · get_code("3f0fe82035c4dbf7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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
import numpy as np

torch.manual_seed(0)

# Create prediction tensor with log probabilities (since function applies np.exp)
# For binary classification (2 classes), batch size 5
batch_size = 5
num_classes = 2

# Log probabilities (will be exponentiated in the function)
prediction = torch.log(torch.softmax(torch.randn(batch_size, num_classes), dim=1))

# Target labels (integers 0 or 1 for binary classification)
target = torch.tensor([0, 1, 0, 1, 1], dtype=torch.long).unsqueeze(1)

ARGS = {
    'prediction': prediction,
    'target': target
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy, sklearn, 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
collate_pool Ran
ran; our contract draft was wrong, not the code
usccolumbia/tsdnn/tsdnn/data.py
code served (permissive licence) · get_code("3f6dd8d427414044")
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 torch
import numpy as np

torch.manual_seed(0)

# Create sample data for a dataset list
# Each element is a tuple: ((atom_fea, nbr_fea, nbr_fea_idx), target, cif_id)

# First sample: 3 atoms
atom_fea_1 = torch.randn(3, 4)  # 3 atoms, 4 features each
nbr_fea_1 = torch.randn(3, 5, 6)  # 3 atoms, 5 neighbors, 6 neighbor features
nbr_fea_idx_1 = torch.LongTensor([[0, 1, 2, 1, 0],
                                   [1, 2, 0, 1, 2],
                                   [2, 0, 1, 2, 0]])  # 3 atoms, 5 neighbor indices
target_1 = torch.tensor([1.5])
cif_id_1 = "crystal_001"

# Second sample: 2 atoms
atom_fea_2 = torch.randn(2, 4)  # 2 atoms, 4 features each
nbr_fea_2 = torch.randn(2, 5, 6)  # 2 atoms, 5 neighbors, 6 neighbor features
nbr_fea_idx_2 = torch.LongTensor([[0, 1, 0, 1, 0],
                                   [1, 0, 1, 0, 1]])  # 2 atoms, 5 neighbor indices
target_2 = torch.tensor([2.3])
cif_id_2 = "crystal_002"

# Construct dataset_list
dataset_list = [
    ((atom_fea_1, nbr_fea_1, nbr_fea_idx_1), target_1, cif_id_1),
    ((atom_fea_2, nbr_fea_2, nbr_fea_idx_2), target_2, cif_id_2)
]

ARGS = {
    "dataset_list": dataset_list
}
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 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
get_cosine_schedule_with_warmup Ran
ran on a synthesized input
ifsheldon/MPL_Lightning/mpl_lightning/pl_modules.py
code served (permissive licence) · get_code("ddf3bcf79dbb01ce")
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 LambdaLR. 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 torch.nn as nn
from torch.optim import SGD

torch.manual_seed(0)

# Create a small model
model = nn.Linear(4, 3)

# Create an optimizer
optimizer = SGD(model.parameters(), lr=0.1)

# Define arguments
ARGS = {
    "optimizer": optimizer,
    "num_warmup_steps": 100,
    "num_training_steps": 1000,
    "num_wait_steps": 0,
    "num_cycles": 0.5,
    "last_epoch": -1,
}
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 LambdaLR — a LambdaLR output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a LambdaLR 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_pos_unl_val_test_loader Ran
ran on a synthesized input
usccolumbia/tsdnn/tsdnn/data.py
code served (permissive licence) · get_code("ae57990ad886dc45")
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
import random
from torch.utils.data import Dataset, DataLoader
from torch.utils.data.dataloader import default_collate

torch.manual_seed(0)

class SimpleDataset(Dataset):
    def __init__(self, size=100, pos_len=None):
        self.size = size
        self.pos_len = pos_len if pos_len is not None else size // 2
        self.data = torch.randn(size, 4)
        self.labels = torch.randint(0, 2, (size,))
    
    def __len__(self):
        return self.size
    
    def __getitem__(self, idx):
        return self.data[idx], self.labels[idx]

class SimpleDatasetWithPosLen(Dataset):
    def __init__(self, size=100, pos_len=30):
        self.size = size
        self.pos_len = pos_len
        self.data = torch.randn(size, 4)
        self.labels = torch.randint(0, 2, (size,))
    
    def __len__(self):
        return self.size
    
    def __getitem__(self, idx):
        return self.data[idx], self.labels[idx]

# Create datasets
labeled_ds = SimpleDataset(size=80)
unlabeled_ds = SimpleDataset(size=20)
full_ds = SimpleDatasetWithPosLen(size=100, pos_len=30)

ARGS = {
    'full_dataset': full_ds,
    'labeled_dataset': labeled_ds,
    'unlabeled_dataset': unlabeled_ds,
    'collate_fn': default_collate,
    'batch_size': 16,
    'train_ratio': 0.7,
    'val_ratio': 0.15,
    'test_ratio': 0.15,
    'return_test': True,
    'num_workers': 0,
    'pin_memory': False,
    'uds': False,
    'train_size': None,
    'test_size': None,
    'val_size': None,
}
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
get_train_val_test_loader Ran
ran; our contract draft was wrong, not the code
usccolumbia/tsdnn/pu_cgcnn/cgcnn/data.py
code served (permissive licence) · get_code("3b412198c15a2ba3")
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
from torch.utils.data import Dataset, DataLoader
from torch.utils.data.dataloader import default_collate

torch.manual_seed(0)

class SimpleDataset(Dataset):
    def __init__(self, num_samples=100, feature_dim=4):
        self.num_samples = num_samples
        self.feature_dim = feature_dim
        self.data = torch.randn(num_samples, feature_dim)
        self.labels = torch.randint(0, 2, (num_samples,))
    
    def __len__(self):
        return self.num_samples
    
    def __getitem__(self, idx):
        return self.data[idx], self.labels[idx]

dataset = SimpleDataset(num_samples=100, feature_dim=4)

ARGS = {
    'dataset': dataset,
    'collate_fn': default_collate,
    'batch_size': 16,
    'train_ratio': 0.8,
    'val_ratio': 0.1,
    'test_ratio': 0.1,
    'return_test': True,
    'num_workers': 0,
    'pin_memory': False,
    'train_size': None,
    'test_size': None,
    'val_size': None,
}
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
mae Ran
ran; our contract draft was wrong, not the code
usccolumbia/tsdnn/predict.py
code served (permissive licence) · get_code("8d28ce13cc82e55f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned Tensor, 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)

N = 2

prediction = torch.randn(N, 1)
target = torch.randn(N, 1)

ARGS = {
    'prediction': prediction,
    'target': target
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, 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
x_u_split Ran
ran on a synthesized input
ifsheldon/MPL_Lightning/mpl_lightning/datamodules.py
code served (permissive licence) · get_code("78c331f531a14ce5")
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 math
import numpy as np

# Set random seed for reproducibility
np.random.seed(0)

# Define arguments for x_u_split function
ARGS = {
    "labels": [0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3],  # 12 labels, 4 classes, 3 per class
    "num_labeled": 8,  # 2 samples per class
    "num_classes": 4,
    "expand_labels": True,
    "batch_size": 4,
    "eval_step": 2
}
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
validate Not yet run
unverified (harvested, no recorded run)
usccolumbia/tsdnn/predict.py
code served (permissive licence) · get_code("431b44daf2725a69")
How it ran

no invocation recorded.

validate Not yet run
unverified (harvested, no recorded run)
usccolumbia/tsdnn/pu_cgcnn/cgcnn/predict.py
code served (permissive licence) · get_code("8d0f8832124fc89d")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 7 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 7 later) and 7 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
AutoContrast 19 other papers
1 dated earlier, 18 later
2001.07685 (2020-01)
FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
2104.07713 (2021-04)
Contrastive Learning with Stronger Augmentations
2105.14148 (2021-05)
OpenMatch: Open-set Consistency Regularization for Semi-supervised Learning with Outliers
and 16 more not named here
Brightness 19 other papers
1 dated earlier, 18 later
2001.07685 (2020-01)
FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
2104.07713 (2021-04)
Contrastive Learning with Stronger Augmentations
2105.14148 (2021-05)
OpenMatch: Open-set Consistency Regularization for Semi-supervised Learning with Outliers
and 16 more not named here
Color 18 other papers
1 dated earlier, 17 later
2001.07685 (2020-01)
FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
2104.07713 (2021-04)
Contrastive Learning with Stronger Augmentations
2105.14148 (2021-05)
OpenMatch: Open-set Consistency Regularization for Semi-supervised Learning with Outliers
and 15 more not named here
mae 5 other papers
1 dated earlier, 4 later
1710.10324 (2017-10)
Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
2206.04968 (2022-06)
What Information is Necessary and Sufficient to Predict Materials Properties using Machine Learning?
2401.00096 (2024-01)
A foundation model for atomistic materials chemistry
and 2 more not named here
class_eval 2 other papers
1 dated earlier, 1 later
1710.10324 (2017-10)
Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
2401.00096 (2024-01)
A foundation model for atomistic materials chemistry
collate_pool 2 other papers
2 later
2104.10869 (2021-04)
CrysXPP:An Explainable Property Predictor for Crystalline Materials
2602.04075 (2026-02)
Thermodynamic assessment of machine learning models for solid-state synthesis prediction
get_train_val_test_loader 1 other paper
1 later
2602.04075 (2026-02)
Thermodynamic assessment of machine learning models for solid-state synthesis prediction

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 2026-11-21.

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 present Meta Pseudo Labels, a semi-supervised learning method that achieves a new state-of-the-art top-1 accuracy of 90.2% on ImageNet, which is 1.6% better than the existing state-of-the-art. Like Pseudo Labels, Meta Pseudo Labels has a teacher network to generate pseudo labels on unlabeled data to teach a student network. However, unlike Pseudo Labels where the teacher is fixed, the teacher in Meta Pseudo Labels is constantly adapted by the feedback of the student's performance on the labeled dataset. As a result, the teacher generates better pseudo labels to teach the student. Our code will be available at https://github.com/google-research/google-research/tree/master/meta_pseudo_labels.

For agents

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

get_harvested_code_for_paper("2003.10580")
get_code_for_paper("2003.10580")
have("2003.10580")

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

curl https://syntology.ai/api/ran/2003.10580.json

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

[![Syntology run record](https://syntology.ai/api/ran/2003.10580.svg)](https://syntology.ai/paper/2003.10580)

Connect an agent — have() is free.