SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.15933 · ACL · 2023

A Survey on Asking Clarification Questions Datasets in Conversational Systems

Emine Yilmaz, Qiang Zhang, Xi Wang, Aldo Lipani, Yue Feng, Hossein Rahmani

arXiv · PDF · Open in the Atlas

Code that ran

14 of 16 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 14 have no contract verdict; 0 failures of our instrument; 2 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/2305.15933.json · badge: /api/ran/2305.15933.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
msra-nlc/MSParS_V2.0 canonical (classifier label)
classifier basis: census
11 ran · 2 no recorded run
aliannejadi/qulac canonical (classifier label)
classifier basis: census
3 ran · 0 no recorded run
FunctionStatusWhere it lives
calculate_accuracy Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/p1_HierarchicalAttention_train.py
code served (permissive licence) · get_code("283c08752ab07dbf")
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 numpy as np

torch.manual_seed(0)

# Analyzing the function:
# - labels_predicted: iterable of label indices (used in for loop)
# - labels: iterable of numeric values (converted to list, indexed, compared with > 0)
# - eval_counter: integer (compared with < 2)

# The function checks which indices in labels have label > 0,
# then counts how many of labels_predicted are in that set.
# However, there's a bug in the code (flag is set in loop but checked outside),
# but we'll provide valid inputs anyway.

# Create labels with some positive and some non-positive values
labels = torch.tensor([0, 1, 2, 0, 3], dtype=torch.float32)

# Create predicted labels as indices (0-4 range)
labels_predicted = [1, 3, 4]

# eval_counter as an integer
eval_counter = 0

ARGS = {
    'labels_predicted': labels_predicted,
    'labels': labels,
    'eval_counter': eval_counter
}
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 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
calculate_accuracy Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a09_DynamicMemoryNet/a8_train.py
code served (permissive licence) · get_code("9b8a7f0a4be941c7")
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 numpy as np

torch.manual_seed(0)

# Analyze the function:
# - labels_predicted: iterable that gets looped over, elements used as dict keys
# - labels: converted to list, elements compared with > 0
# - eval_counter: compared with < 2, used in conditional print

# The function has a bug (flag is assigned but count logic is outside the loop),
# but we construct valid inputs anyway.

labels_predicted = [0, 2]  # List of predicted label indices
labels = torch.tensor([0, 1, 2, 1])  # Tensor with some values > 0
eval_counter = 0  # Integer counter

ARGS = {
    "labels_predicted": labels_predicted,
    "labels": labels,
    "eval_counter": eval_counter
}
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 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
compute_single_label Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a02_TextCNN/p7_TextCNN_model.py
code served (permissive licence) · get_code("40d0f2232c402e7c")
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 int; 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)

# Analyze the function: compute_single_label takes a list and processes each element
# by summing the previous, current, and next elements, then thresholding at 2
# It returns a list of binary values (0 or 1)

# Create a simple list of integers as input
listt = [0, 1, 1, 0, 1, 2]

ARGS = {
    "listt": listt
}
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 list, ndim 1, dtype family int — 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
decode_id Ran
ran on a synthesized input
aliannejadi/qulac/src/utils/utils.py
code served (permissive licence) · get_code("5d30e0bd6048f033")
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

def decode_id(rec_id):
    tokens = rec_id.split('-')
    topic_id, facet_id, question_id, with_ans = (tokens[0], tokens[1], tokens[-2], tokens[-1])
    hisotry_ids = ()
    for k in range(2, 2 + len(tokens) - 4):
        hisotry_ids += (tokens[k],)
    return (topic_id, facet_id, hisotry_ids, question_id, with_ans)

# Build a valid rec_id string
# Format: topic-facet-history1-history2-...-question-with_ans
# Minimum format with no history: topic-facet-question-with_ans (4 tokens)
# With history items: topic-facet-hist1-hist2-...-question-with_ans

rec_id = "topic1-facet2-history1-history2-question3-answer_yes"

ARGS = {
    "rec_id": rec_id
}
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
encode_id Ran
ran on a synthesized input
aliannejadi/qulac/src/utils/utils.py
code served (permissive licence) · get_code("8ef525178610a9a3")
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

torch.manual_seed(0)

ARGS = {
    'topic_id': 1,
    'facet_id': 2,
    'hisotry_ids': [10, 20, 30],
    'question_id': 42,
    'with_ans': True
}
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
extract_argmax_and_embed Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/p1_seq2seq.py
code served (permissive licence) · get_code("3a19987885a91fe3")
How it ran

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

import tensorflow as tf
import numpy as np

# Create a small embedding tensor
# Shape: (vocab_size=5, embedding_dim=4)
embedding = tf.constant(np.random.randn(5, 4).astype(np.float32))

# Create output projection: (W, B)
# W shape: (4, 3) - projects from embedding_dim to num_classes
# B shape: (3,) - bias for num_classes
W = tf.constant(np.random.randn(4, 3).astype(np.float32))
B = tf.constant(np.random.randn(3).astype(np.float32))
output_projection = (W, B)

ARGS = {
    "embedding": embedding,
    "output_projection": output_projection,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_input_y Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/p1_HierarchicalAttention_model_transformer.py
code served (permissive licence) · get_code("413b9f01ceeb3a80")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family int, shape (2,) 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

# Analyze the function:
# - i: parameter used but not referenced in function body
# - input_x: needs to be indexable by batch_size, each element needs np.sum()
# - batch_size: integer specifying number of batches

# Create a small batch of input data
batch_size = 2
# Create input_x with 2 samples, each with 4 features
# Sample 0: sum = 5 (< 10) -> label 0
# Sample 1: sum = 15 (> 10) -> label 2
input_x = np.array([
    [1.0, 2.0, 1.0, 1.0],  # sum = 5
    [5.0, 4.0, 3.0, 3.0]   # sum = 15
])

i = 0  # parameter used but not in function body

ARGS = {
    "i": i,
    "input_x": input_x,
    "batch_size": batch_size
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family int, shape (2,) — shape (2,) 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_label_using_logits Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/p1_HierarchicalAttention_predict.py
code served (permissive licence) · get_code("9636df81bbc3414f")
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 numpy as np

# Create a small vocabulary index to word label mapping
vocabulary_index2word_label = {
    0: "cat",
    1: "dog",
    2: "bird",
    3: "fish",
    4: "elephant"
}

# Create small logits array (5 elements to match vocabulary size)
logits = np.array([0.1, 0.5, 2.3, 1.1, 0.8])

ARGS = {
    "logits": logits,
    "vocabulary_index2word_label": vocabulary_index2word_label,
    "top_number": 5
}
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 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
get_label_using_logits Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/p1_HierarchicalAttention_train.py
code served (permissive licence) · get_code("9167bcc3b577f7ee")
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 int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create a small logits array
# The function calls np.argmax(logits), so we need a valid numpy array
logits = np.array([0.1, 0.5, 0.2, 0.15])

ARGS = {
    "logits": logits,
    "top_number": 5
}
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 list, ndim 1, dtype family int — 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
get_label_using_logits Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a09_DynamicMemoryNet/a8_train.py
code served (permissive licence) · get_code("425e506187f1c317")
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 int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create sample logits with shape (batch_size=2, num_classes=4)
logits = np.array([
    [0.1, 0.5, 0.3, 0.1],
    [0.2, 0.1, 0.6, 0.1]
], dtype=np.float32)

ARGS = {
    "logits": logits,
    "top_number": 5
}
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 list, ndim 1, dtype family int — 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
get_label_using_logits_with_value Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/p1_HierarchicalAttention_predict.py
code served (permissive licence) · get_code("13a27928f135f31f")
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 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create logits as a 1D numpy array
logits = np.array([0.1, 0.5, 0.3, 0.9, 0.2, 0.7])

# Create vocabulary_index2word_label as a dict mapping indices to words
vocabulary_index2word_label = {
    0: "label_0",
    1: "label_1",
    2: "label_2",
    3: "label_3",
    4: "label_4",
    5: "label_5"
}

# Set top_number parameter
top_number = 5

ARGS = {
    "logits": logits,
    "vocabulary_index2word_label": vocabulary_index2word_label,
    "top_number": top_number
}
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, ndim 2, 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_label_y Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a02_TextCNN/p7_TextCNN_model.py
code served (permissive licence) · get_code("f1a16c0dadccbcf8")
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 (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create a small test array with shape (2, 4)
# The function expects input_x to be a numpy array with .shape attribute
# It iterates over rows and applies compute_single_label to each row
input_x = np.array([
    [1, 0, 1, 0],
    [0, 1, 1, 1]
], dtype=np.float32)

ARGS = {
    "input_x": input_x
}
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 (2, 4) — shape (2, 4) 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
remove_symbols Ran
ran on a synthesized input
aliannejadi/qulac/src/utils/utils.py
code served (permissive licence) · get_code("1fb9df9e0cbc7c7b")
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

def remove_symbols(input_str):
    input_str = input_str.replace('\\', '')
    return input_str

ARGS = {
    'input_str': 'hello\\world\\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 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
transform_multilabel_as_multihot Ran
ran on a synthesized input
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/data_util.py
code served (permissive licence) · get_code("aa03d525f10278ed")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (10,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Construct arguments for transform_multilabel_as_multihot
label_list = [0, 1, 4]
label_size = 10

ARGS = {
    "label_list": label_list,
    "label_size": label_size
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (10,) — shape (10,) 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
create_vocabulary Not yet run
unverified (harvested, no recorded run)
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/data_util.py
code served (permissive licence) · get_code("f8f1d8c68d7c71c2")
How it ran

no invocation recorded.

rnn_decoder_with_attention Not yet run
unverified (harvested, no recorded run)
msra-nlc/MSParS_V2.0/code/classification/a05_HierarchicalAttentionNetwork/p1_seq2seq.py
code served (permissive licence) · get_code("7fd5c811b5950db3")
How it ran

no invocation recorded.

Also present in other papers' code

Of 16 samples lifted from this paper's linked repositories, 16 carry a function identity and 0 do not. Of the 16 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 16 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 ability to understand a user's underlying needs is critical for conversational systems, especially with limited input from users in a conversation. Thus, in such a domain, Asking Clarification Questions (ACQs) to reveal users' true intent from their queries or utterances arise as an essential task. However, it is noticeable that a key limitation of the existing ACQs studies is their incomparability, from inconsistent use of data, distinct experimental setups and evaluation strategies. Therefore, in this paper, to assist the development of ACQs techniques, we comprehensively analyse the current ACQs research status, which offers a detailed comparison of publicly available datasets, and discusses the applied evaluation metrics, joined with benchmarks for multiple ACQs-related tasks. In particular, given a thorough analysis of the ACQs task, we discuss a number of corresponding research directions for the investigation of ACQs as well as the development of conversational systems.

For agents

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

get_harvested_code_for_paper("2305.15933")
get_code_for_paper("2305.15933")
have("2305.15933")

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

curl https://syntology.ai/api/ran/2305.15933.json

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

[![Syntology run record](https://syntology.ai/api/ran/2305.15933.svg)](https://syntology.ai/paper/2305.15933)

Connect an agent — have() is free.