SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.09093 · EMNLP · 2023

GLEN: General-Purpose Event Detection for Thousands of Types

Jiawei Han, Heng Ji, Sha Li, Qiusi Zhan, Martha Palmer, Kathryn Conger

arXiv · PDF · Open in the Atlas

Code that ran

12 of 15 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 12 have no contract verdict; 0 failures 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/2303.09093.json · badge: /api/ran/2303.09093.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
ZQS1943/GLEN canonical (classifier label)
classifier basis not recorded on the link
12 ran · 3 no recorded run
FunctionStatusWhere it lives
collate_fn_TC_sen Ran
ran on a synthesized input
ZQS1943/GLEN/model/dataset.py
code served (permissive licence) · get_code("e8a22e3d63e08e3e")
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

torch.manual_seed(0)

# Create a batch of sample data items
# Each item is a dict with keys: 'id', 'event_idx', 'event_id', 'input_ids', 'mask_token_mask'
batch = [
    {
        'id': 'doc_0',
        'event_idx': 0,
        'event_id': 'evt_001',
        'input_ids': [101, 2054, 2003, 102],
        'mask_token_mask': [False, True, False, False]
    },
    {
        'id': 'doc_1',
        'event_idx': 1,
        'event_id': 'evt_002',
        'input_ids': [101, 2003, 2054, 102],
        'mask_token_mask': [False, False, True, False]
    }
]

ARGS = {
    'batch': batch
}
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
collate_fn_TI_sen Ran
ran on a synthesized input
ZQS1943/GLEN/model/dataset.py
code served (permissive licence) · get_code("8475e39d13821a6d")
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

torch.manual_seed(0)

# Create a minimal batch of data
# Each item in the batch needs 'index' and 'tokenized_text_ids' keys
batch = [
    {
        'index': 0,
        'tokenized_text_ids': [101, 2054, 2003, 102]
    },
    {
        'index': 1,
        'tokenized_text_ids': [101, 3231, 2054, 102]
    }
]

ARGS = {
    'batch': batch
}
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
collate_fn_TR_sen Ran
ran on a synthesized input
ZQS1943/GLEN/model/dataset.py
code served (permissive licence) · get_code("9be8a0b5e69c13f1")
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

torch.manual_seed(0)

# The function expects a batch of dicts, each containing:
# - 'index': an integer
# - 'tokenized_text_ids': a list/sequence of token ids

batch = [
    {
        'index': 0,
        'tokenized_text_ids': [101, 2054, 2003, 102]
    },
    {
        'index': 1,
        'tokenized_text_ids': [101, 2054, 2054, 102]
    }
]

ARGS = {
    'batch': batch
}
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
flatten_and_batch_shift_indices Ran
ran on a synthesized input
ZQS1943/GLEN/model/allennlp_span_utils.py
code served (permissive licence) · get_code("38ff6aab7932baab")
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 Tensor, ndim 1, dtype family int, shape (6,) as a later run recorded it. 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, 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)

# Create indices tensor with shape (batch_size=2, d_1=3)
# Values must be in valid range [0, sequence_length-1]
indices = torch.tensor([
    [0, 1, 2],
    [3, 4, 5]
], dtype=torch.long)

# sequence_length must be greater than max index value (5)
sequence_length = 10

ARGS = {
    "indices": indices,
    "sequence_length": sequence_length
}
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 Tensor, ndim 1, dtype family int, shape (6,) — shape (6,) 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_context_representation_multiple_mentions_idxs Ran
ran on a synthesized input
ZQS1943/GLEN/model/data_process.py
code served (permissive licence) · get_code("4c689d496eb7575b")
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

class MockTokenizer:
    def convert_ids_to_tokens(self, ids):
        return [f"token_{i}" for i in ids]

# Create sample data
sample = {
    'tokenized_trigger_idxs': [[2, 4], [6, 8]],
    'tokenized_text_ids': [101, 10, 20, 30, 40, 50, 60, 70, 80, 90, 102],
    'mentions': ['mention1', 'mention2'],
    'label_ids': [1, 2],
    'triggers': ['trigger1', 'trigger2'],
    'xpo_ids': [10, 20],
    'xpo_titles': ['title1', 'title2'],
    'label': ['label1', 'label2'],
    'rolesets': ['roleset1', 'roleset2'],
    'true_label': ['true1', 'true2']
}

tokenizer = MockTokenizer()

ARGS = {
    'sample': sample,
    'tokenizer': tokenizer,
    'max_seq_length': 128,
    'add_sent_event_token': False,
    'sent_tag_id': 1
}
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 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_device_of Ran
ran on a synthesized input
ZQS1943/GLEN/model/allennlp_span_utils.py
code served (permissive licence) · get_code("577de867e8601b82")
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 int, ndim 0, dtype family int, 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)

# Create a CPU tensor
cpu_tensor = torch.randn(2, 4)

ARGS = {
    "tensor": cpu_tensor
}
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 int, ndim 0, dtype family int, 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
get_model_obj Ran
ran on a synthesized input
ZQS1943/GLEN/model/encoder.py
code served (permissive licence) · get_code("ed37d2c4c5923196")
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 SimpleModel. 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

torch.manual_seed(0)

# The function checks if model has a 'module' attribute
# If it does, it returns model.module, otherwise returns model
# So we need to test both cases

# Create a simple nn.Module
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Test case 1: model without 'module' attribute
simple_model = SimpleModel()

# Test case 2: model wrapped with 'module' attribute (e.g., DataParallel)
wrapped_model = nn.Module()
wrapped_model.module = SimpleModel()

# We'll use the wrapped model since it exercises the hasattr branch
model_to_test = wrapped_model

ARGS = {
    "model": model_to_test
}
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 SimpleModel — a SimpleModel output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModel 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_optimizer Ran
ran on a synthesized input
ZQS1943/GLEN/model/optimizer.py
code served (permissive licence) · get_code("a332aeb0adc8f092")
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 torch.nn as nn
from transformers import get_linear_schedule_with_warmup

torch.manual_seed(0)

# Create a minimal model with named parameters
class MinimalModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.layer_norm = nn.LayerNorm(3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.layer_norm(x)
        x = self.linear2(x)
        return x

model = MinimalModel()

# Total number of training steps
t_total = 100

# Arguments dictionary with required keys
ARGS = {
    'model': model,
    't_total': t_total,
    'args': {
        'learning_rate': 5e-5,
        'warmup_steps': 10
    }
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — 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_range_vector Ran
ran on a synthesized input
ZQS1943/GLEN/model/allennlp_span_utils.py
code served (permissive licence) · get_code("8b5bf32ce9d8e16c")
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 Tensor, ndim 1, dtype family int, shape (5,) as a later run recorded it. 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, 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)

# The function takes:
# - size: int - the desired size of the range vector
# - device: int - the device index (-1 for CPU, >=0 for CUDA)

# For CPU-only execution, we pass device=-1 to avoid CUDA calls
# We test with a small size value

ARGS = {
    "size": 5,
    "device": -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 Tensor, ndim 1, dtype family int, shape (5,) — shape (5,) 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
loss_for_type_classification Ran
ran on a synthesized input
ZQS1943/GLEN/model/encoder.py
code served (permissive licence) · get_code("c2e8a2dd778aadba")
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 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)

# Create tensors for yes_scores and cand_mask
# yes_scores should be a 1D tensor of scores
# cand_mask should be a boolean mask of the same size
yes_scores = torch.tensor([0.5, 0.8, 0.2, 0.9], dtype=torch.float32)
cand_mask = torch.tensor([True, False, True, False], dtype=torch.bool)

device = torch.device('cpu')
th = 0.1

ARGS = {
    'yes_scores': yes_scores,
    'cand_mask': cand_mask,
    'device': device,
    'th': th
}
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 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
loss_for_type_ranking Ran
ran on a synthesized input
ZQS1943/GLEN/model/encoder.py
code served (permissive licence) · get_code("f767a4a3eb97ee72")
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 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)

# Analyze function requirements:
# - scores: tensor, indexed by [batch_idx], then by indices in negative_samples and candidate_label_sets
# - candidate_label_sets: list of lists of indices for each batch item
# - negative_samples: list of index lists for each batch item
# - device: torch device
# - th: float threshold

# Build tiny example with batch_size=2
batch_size = 2
num_labels = 5  # number of possible labels to score

# scores: shape (2, 5) - batch of 2, each with 5 label scores
scores = torch.randn(batch_size, num_labels)

# candidate_label_sets: list of batch_size, each containing lists of indices
# For batch item 0: candidate sets are [[0, 1], [2]]
# For batch item 1: candidate sets are [[1, 3]]
candidate_label_sets = [
    [[0, 1], [2]],  # batch item 0: 2 candidate sets
    [[1, 3]],       # batch item 1: 1 candidate set
]

# negative_samples: list of batch_size, each containing a list of negative indices
negative_samples = [
    [3, 4],    # batch item 0: negative indices
    [0, 2, 4], # batch item 1: negative indices
]

device = torch.device('cpu')
th = 1.0

ARGS = {
    'scores': scores,
    'candidate_label_sets': candidate_label_sets,
    'negative_smaples': negative_samples,
    'device': device,
    'th': th,
}
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 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
sort_mentions Ran
ran on a synthesized input
ZQS1943/GLEN/model/data_process.py
code served (permissive licence) · get_code("528d054aece22a60")
How it ran

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

import torch

torch.manual_seed(0)

# Example list with 5 mentions (could be strings, dicts, or any objects)
lst = ["mention_0", "mention_1", "mention_2", "mention_3", "mention_4"]

# sort_map defines where each element at index i should go in the new list
# For example: original index 0 goes to position 2, index 1 goes to position 0, etc.
sort_map = {0: 2, 1: 0, 2: 4, 3: 1, 4: 3}

ARGS = {
    "lst": lst,
    "sort_map": sort_map
}
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, 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
evaluate Not yet run
unverified (harvested, no recorded run)
ZQS1943/GLEN/model/predict_type_classifier.py
code served (permissive licence) · get_code("e1211212d3a27669")
How it ran

no invocation recorded.

predict_top_k Not yet run
unverified (harvested, no recorded run)
ZQS1943/GLEN/model/predict_type_ranking.py
code served (permissive licence) · get_code("37504ee9631718e0")
How it ran

no invocation recorded.

sentence_tokenization Not yet run
unverified (harvested, no recorded run)
ZQS1943/GLEN/model/predict_sentence.py
code served (permissive licence) · get_code("9171f21b673fd36e")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 15 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.

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

The progress of event extraction research has been hindered by the absence of wide-coverage, large-scale datasets. To make event extraction systems more accessible, we build a generalpurpose event detection dataset GLEN which covers 205K event mentions with 3,465 different types, making it more than 20x larger in ontology than today's largest event dataset. GLEN is created by utilizing the DWD Overlay, which provides a mapping between Wikidata Qnodes and PropBank rolesets. This enables us to use the abundant existing annotation for PropBank as distant supervision. In addition, we also propose a new multi-stage event detection model CEDAR specifically designed to handle the large ontology size in GLEN. We show that our model exhibits superior performance compared to a range of baselines including InstructGPT. Finally, we perform error analysis and show that label noise is still the largest challenge for improving performance for this new dataset. 1

For agents

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

get_harvested_code_for_paper("2303.09093")
get_code_for_paper("2303.09093")
have("2303.09093")

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

curl https://syntology.ai/api/ran/2303.09093.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.09093.svg)](https://syntology.ai/paper/2303.09093)

Connect an agent — have() is free.