SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2009.05387 · 2020

IndoNLU: Benchmark and Resources for Evaluating Indonesian Natural Language Understanding

arXiv · PDF · Open in the Atlas

Code that ran

11 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; 1 honoured the contract we drafted, 0 violated it and 9 have no contract verdict; 1 failure of our instrument; 3 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/2009.05387.json · badge: /api/ran/2009.05387.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
indobenchmark/indonlu canonical (classifier label)
classifier basis: census
11 ran · 3 no recorded run
FunctionStatusWhere it lives
emotion_detection_metrics_fn Ran
ran on a synthesized input
indobenchmark/indonlu/utils/metrics.py
code served (permissive licence) · get_code("536eda93d03af310")
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
import numpy as np
from sklearn.metrics import accuracy_score, f1_score, recall_score, precision_score

torch.manual_seed(0)

# The function takes two parameters:
# - list_hyp: predicted emotion labels (list of integers)
# - list_label: true emotion labels (list of integers)
# Both are passed to sklearn metrics functions which expect array-like objects

# Create small sample emotion predictions and labels
# Using 3 emotion classes (0, 1, 2) with 10 samples
list_hyp = [0, 1, 2, 0, 1, 2, 0, 1, 2, 0]
list_label = [0, 1, 1, 0, 1, 2, 0, 2, 2, 0]

ARGS = {
    'list_hyp': list_hyp,
    'list_label': list_label,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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 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
end_of_chunk Ran
ran on a synthesized input
indobenchmark/indonlu/utils/conlleval.py
code served (permissive licence) · get_code("abc247eaad47c595")
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

ARGS = {
    'prev_tag': 'B',
    'tag': 'I',
    'prev_type': 'PER',
    'type_': 'PER'
}
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
forward_sequence_classification Ran
ran on a synthesized input
indobenchmark/indonlu/utils/forward_fn.py
code served (permissive licence) · get_code("b0608daef236473c")
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. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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

torch.manual_seed(0)

# Create a minimal mock model that mimics a transformer-like behavior
class MockSequenceClassificationModel(nn.Module):
    def __init__(self, num_labels=3):
        super().__init__()
        self.num_labels = num_labels
        self.embedding = nn.Embedding(100, 8)
        self.linear = nn.Linear(8, num_labels)
    
    def forward(self, input_ids, attention_mask=None, token_type_ids=None, labels=None):
        # input_ids shape: (batch_size, seq_len)
        embeddings = self.embedding(input_ids)  # (batch_size, seq_len, 8)
        # Simple pooling: take mean across sequence
        pooled = embeddings.mean(dim=1)  # (batch_size, 8)
        logits = self.linear(pooled)  # (batch_size, num_labels)
        
        loss = None
        if labels is not None:
            # Simple cross-entropy loss
            loss_fn = nn.CrossEntropyLoss()
            loss = loss_fn(logits, labels.squeeze(-1))
        
        return (loss, logits)

# Create the model
model = MockSequenceClassificationModel(num_labels=3)

# Create index-to-word mapping
i2w = {0: 'label_0', 1: 'label_1', 2: 'label_2'}

# Create batch data with 3 elements: (subword_batch, mask_batch, label_batch)
# batch_size=2, seq_len=4
subword_batch = [[5, 10, 15, 20], [8, 12, 18, 22]]
mask_batch = [[1.0, 1.0, 1.0, 1.0], [1.0, 1.0, 1.0, 1.0]]
label_batch = [[1], [2]]  # Shape (batch_size, 1)

batch_data = (subword_batch, mask_batch, label_batch)

ARGS = {
    'model': model,
    'batch_data': batch_data,
    'i2w': i2w,
    'is_test': False,
    'device': 'cpu'
}
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 — 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
forward_sequence_multi_classification Ran
ran on a synthesized input
indobenchmark/indonlu/utils/forward_fn.py
code served (permissive licence) · get_code("53f0892c0d642652")
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. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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

torch.manual_seed(0)

# Define a minimal model that mimics the expected behavior
class MultiClassificationModel(nn.Module):
    def __init__(self, vocab_size=100, num_labels=5, hidden_size=8):
        super().__init__()
        self.embedding = nn.Embedding(vocab_size, hidden_size)
        self.classifier = nn.Linear(hidden_size, num_labels)
        self.loss_fn = nn.CrossEntropyLoss()
    
    def forward(self, input_ids, attention_mask=None, token_type_ids=None, labels=None):
        # input_ids: (batch_size, seq_len)
        embedded = self.embedding(input_ids)  # (batch_size, seq_len, hidden_size)
        
        # Apply attention mask if provided
        if attention_mask is not None:
            embedded = embedded * attention_mask.unsqueeze(-1)
        
        # Compute logits for each token position
        logits = self.classifier(embedded)  # (batch_size, seq_len, num_labels)
        
        # Compute loss if labels provided
        loss = None
        if labels is not None:
            loss = self.loss_fn(logits.view(-1, logits.size(-1)), labels.view(-1))
        
        return (loss, logits)


# Create the model
model = MultiClassificationModel(vocab_size=100, num_labels=5, hidden_size=8)
model.eval()

# Create batch data (3-tuple format: subword_batch, mask_batch, label_batch)
batch_size = 2
seq_len = 4
num_labels = 5

subword_batch = [[1, 2, 3, 4], [5, 6, 7, 8]]  # (batch_size, seq_len)
mask_batch = [[1.0, 1.0, 1.0, 1.0], [1.0, 1.0, 1.0, 0.0]]  # (batch_size, seq_len)
label_batch = [[0, 1, 2, 3], [1, 2, 3, 4]]  # (batch_size, seq_len)

batch_data = [subword_batch, mask_batch, label_batch]

# Create index-to-word mapping
i2w = {0: "class_0", 1: "class_1", 2: "class_2", 3: "class_3", 4: "class_4"}

ARGS = {
    'model': model,
    'batch_data': batch_data,
    'i2w': i2w,
    'is_test': False,
    'device': 'cpu'
}
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 — 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
forward_word_classification Ran
ran on a synthesized input
indobenchmark/indonlu/utils/forward_fn.py
code served (permissive licence) · get_code("0e01586952d91f1a")
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. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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

torch.manual_seed(0)

# Build a minimal model that satisfies the function's usage
class MinimalWordClassifier(nn.Module):
    def __init__(self, vocab_size=100, num_labels=5, hidden_size=8):
        super().__init__()
        self.embedding = nn.Embedding(vocab_size, hidden_size)
        self.classifier = nn.Linear(hidden_size, num_labels)
    
    def forward(self, subword_batch, subword_to_word_indices_batch, attention_mask=None, token_type_ids=None, labels=None):
        # subword_batch: (batch_size, seq_len)
        embedded = self.embedding(subword_batch)  # (batch_size, seq_len, hidden_size)
        logits = self.classifier(embedded)  # (batch_size, seq_len, num_labels)
        
        loss = None
        if labels is not None:
            loss_fn = nn.CrossEntropyLoss(ignore_index=-100)
            loss = loss_fn(logits.view(-1, logits.size(-1)), labels.view(-1))
        
        return (loss, logits)

model = MinimalWordClassifier(vocab_size=100, num_labels=5, hidden_size=8)

# Build batch_data with 4 elements (no token_type_batch)
batch_size = 2
seq_len = 3
num_labels = 5

subword_batch = [[1, 2, 3], [4, 5, 6]]
mask_batch = [[1.0, 1.0, 1.0], [1.0, 1.0, 1.0]]
subword_to_word_indices_batch = [[0, 1, 2], [0, 1, 2]]
label_batch = [[0, 1, 2], [3, 4, -100]]

batch_data = (subword_batch, mask_batch, subword_to_word_indices_batch, label_batch)

# Build i2w (index to word mapping)
i2w = {0: "word0", 1: "word1", 2: "word2", 3: "word3", 4: "word4", 5: "word5"}

ARGS = {
    "model": model,
    "batch_data": batch_data,
    "i2w": i2w,
    "is_test": False,
    "device": "cpu"
}
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 — 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
gen_embeddings Ran
ran on a synthesized input
indobenchmark/indonlu/utils/functions.py
code served (permissive licence) · get_code("931c15ea00fd6e21")
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 (5, 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
import tempfile
import os

# Create a temporary embedding file
temp_dir = tempfile.mkdtemp()
emb_file_path = os.path.join(temp_dir, "embeddings.txt")

# Write a sample embedding file
with open(emb_file_path, 'w') as f:
    # First line: vocabulary size and embedding dimension
    f.write("5 3\n")
    # Sample word embeddings
    f.write("hello 0.1 0.2 0.3\n")
    f.write("world 0.4 0.5 0.6\n")
    f.write("test 0.7 0.8 0.9\n")

# Create a vocabulary list
vocab_list = ["hello", "world", "test", "unknown", "word"]

ARGS = {
    "vocab_list": vocab_list,
    "emb_path": emb_file_path,
    "emb_dim": None
}
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 (5, 3) — shape (5, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_lr Ran
ran, honoured the contract we drafted
indobenchmark/indonlu/predict.py
code served (permissive licence) · get_code("e9c7d263fd12f86d")
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 torch
import torch.nn as nn
from torch.optim import SGD

torch.manual_seed(0)

# Create a minimal model with parameters
model = nn.Linear(4, 3)

# Create an optimizer with a specific learning rate
learning_rate = 0.001
optimizer = SGD(model.parameters(), lr=learning_rate)

# Create minimal args object (the function only iterates over optimizer.param_groups)
# args is not actually used in the function body, so we can use a simple dict or object
class MinimalArgs:
    pass

args = MinimalArgs()

ARGS = {
    'args': args,
    'optimizer': optimizer
}
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 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
load_vocab Ran
ran on a synthesized input
indobenchmark/indonlu/utils/functions.py
code served (permissive licence) · get_code("4384e3778f7c4a96")
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 tempfile
import os

# Create a temporary file with sample vocabulary
temp_dir = tempfile.mkdtemp()
vocab_path = os.path.join(temp_dir, 'vocab.txt')

# Write sample vocabulary to file
with open(vocab_path, 'w') as f:
    f.write('apple\n')
    f.write('banana\n')
    f.write('cherry\n')
    f.write('date\n')

ARGS = {
    'path': vocab_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 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
metrics_to_string Ran
ran; our contract draft was wrong, not the code
indobenchmark/indonlu/predict.py
code served (permissive licence) · get_code("2c8f98e551054b67")
How it ran

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

import torch
import numpy as np

torch.manual_seed(0)

# The function metrics_to_string takes a metric_dict and iterates over its items,
# formatting each key-value pair where value is a float.
# It requires a dictionary with string keys and numeric values (floats).

metric_dict = {
    'loss': 2.5432,
    'accuracy': 0.9187,
    'f1_score': 0.8734,
    'precision': 0.9045
}

ARGS = {
    'metric_dict': metric_dict
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
parse_tag Ran
ran on a synthesized input
indobenchmark/indonlu/utils/conlleval.py
code served (permissive licence) · get_code("2e9480de024f3eec")
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 re

def parse_tag(t):
    m = re.match('^([^-]*)-(.*)$', t)
    return m.groups() if m else (t, '')

ARGS = {
    't': 'model-v1'
}
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 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
start_of_chunk Ran
ran on a synthesized input
indobenchmark/indonlu/utils/conlleval.py
code served (permissive licence) · get_code("8a646b990ee217c9")
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.

# Setup for start_of_chunk function
# This function checks if a new chunk starts based on BIO/BIOES tags and types

ARGS = {
    "prev_tag": "O",
    "tag": "B",
    "prev_type": "PER",
    "type_": "PER"
}
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
aspect_extraction_metrics_fn Not yet run
unverified (harvested, no recorded run)
indobenchmark/indonlu/utils/metrics.py
code served (permissive licence) · get_code("0fb5de45230b147d")
How it ran

no invocation recorded.

ner_metrics_fn Not yet run
unverified (harvested, no recorded run)
indobenchmark/indonlu/utils/metrics.py
code served (permissive licence) · get_code("b9229e6df7796a72")
How it ran

no invocation recorded.

predict Not yet run
unverified (harvested, no recorded run)
indobenchmark/indonlu/predict.py
code served (permissive licence) · get_code("da384f25986bd37e")
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, 3 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 2 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
parse_tag 3 other papers
3 dated earlier
1805.10190 (2018-05)
Snips Voice Platform: an embedded Spoken Language Understanding system for private-by-design voice interfaces
2004.11727 (2020-04)
Coach: A Coarse-to-Fine Approach for Cross-domain Slot Filling
2006.01563 (2020-06)
Exploring Cross-sentence Contexts for Named Entity Recognition with BERT
get_lr 2 other papers
2 later
2109.07684 (2021-09)
Language Models are Few-shot Multilingual Learners
2309.10661 (2023-09)
NusaWrites: Constructing High-Quality Corpora for Underrepresented and Extremely Low-Resource Languages
metrics_to_string 2 other papers
2 later
2109.07684 (2021-09)
Language Models are Few-shot Multilingual Learners
2309.10661 (2023-09)
NusaWrites: Constructing High-Quality Corpora for Underrepresented and Extremely Low-Resource Languages

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

Although Indonesian is known to be the fourth most frequently used language over the internet, the research progress on this language in the natural language processing (NLP) is slow-moving due to a lack of available resources. In response, we introduce the first-ever vast resource for the training, evaluating, and benchmarking on Indonesian natural language understanding (IndoNLU) tasks. IndoNLU includes twelve tasks, ranging from single sentence classification to pair-sentences sequence labeling with different levels of complexity. The datasets for the tasks lie in different domains and styles to ensure task diversity. We also provide a set of Indonesian pre-trained models (IndoBERT) trained from a large and clean Indonesian dataset Indo4B collected from publicly available sources such as social media texts, blogs, news, and websites. We release baseline models for all twelve tasks, as well as the framework for benchmark evaluation, and thus it enables everyone to benchmark their system performances.

For agents

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

get_harvested_code_for_paper("2009.05387")
get_code_for_paper("2009.05387")
have("2009.05387")

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

curl https://syntology.ai/api/ran/2009.05387.json

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

[![Syntology run record](https://syntology.ai/api/ran/2009.05387.svg)](https://syntology.ai/paper/2009.05387)

Connect an agent — have() is free.