SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1711.04436 · 2017

SQLNet: Generating Structured Queries From Natural Language Without Reinforcement Learning

arXiv · PDF · Open in the Atlas

Code that ran

11 of 19 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 0 have no contract verdict; 10 failures of our instrument; 8 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/1711.04436.json · badge: /api/ran/1711.04436.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
Baidi96/text2sql community (classifier label)
classifier basis not recorded on the link
4 ran · 2 no recorded run
CX000/sqlnet_inference_py36 community (classifier label)
classifier basis not recorded on the link
3 ran · 4 no recorded run
PriyankaDatar/NLP_project_BIDAF extension (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
PriyankaDatar/NLP_SQL_Project community (classifier label)
classifier basis not recorded on the link
1 ran · 1 no recorded run
PriyankaDatar/NLP_Project_Modfications community (classifier label)
classifier basis not recorded on the link
0 ran · 1 no recorded run
FunctionStatusWhere it lives
att_flow_layer Ran
ran; our contract draft was wrong, not the code
PriyankaDatar/NLP_project_BIDAF/sqlnet/model/modules/net_utils.py
code served (permissive licence) · get_code("d74e5a33a9136d0d")
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) and returned Tensor, ndim 3, dtype family float, shape (2, 3, 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 torch
import torch.nn as nn
import torch.nn.functional as F

torch.manual_seed(0)

# Define small attention weight modules
class AttWeightModule(nn.Module):
    def __init__(self, input_size, output_size):
        super().__init__()
        self.linear = nn.Linear(input_size, output_size)
    
    def forward(self, x):
        return self.linear(x)

class SoftmaxModule(nn.Module):
    def __init__(self):
        super().__init__()
    
    def forward(self, x):
        return F.softmax(x, dim=-1)

# Set up dimensions
batch_size = 2
c_len = 3
q_len = 4
hidden_size = 2

# Create tensors
c = torch.randn(batch_size, c_len, hidden_size * 2)  # (2, 3, 4)
q = torch.randn(batch_size, q_len, hidden_size * 2)  # (2, 4, 4)

# Create attention weight modules
att_weight_cq = AttWeightModule(hidden_size * 2, 1)
att_weight_c = AttWeightModule(hidden_size * 2, 1)
att_weight_q = AttWeightModule(hidden_size * 2, 1)

# Create softmax module
softmax = SoftmaxModule()

ARGS = {
    'c': c,
    'q': q,
    'softmax': softmax,
    'att_weight_cq': att_weight_cq,
    'att_weight_c': att_weight_c,
    'att_weight_q': att_weight_q
}
Setup block drafted by
a language model, Claude Haiku 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 3, dtype family float, shape (2, 3, 4) — shape (2, 3, 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
best_model_name Ran
ran; our contract draft was wrong, not the code
Baidi96/text2sql/sqlnet/utils.py
code served (permissive licence) · get_code("17e87858aaa02d44")
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.

class Args:
    def __init__(self, dataset=0, baseline=False, train_emb=False, rl=False, ca=False):
        self.dataset = dataset
        self.baseline = baseline
        self.train_emb = train_emb
        self.rl = rl
        self.ca = ca

ARGS = {
    'args': Args(dataset=1, baseline=False, train_emb=True, rl=True, ca=True),
    'for_load': False
}
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
col_name_encode Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
PriyankaDatar/NLP_project_BIDAF/sqlnet/model/modules/net_utils.py
code served (permissive licence) · get_code("f2b7fb04ad23ccb2")
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) 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
import numpy as np
from torch.autograd import Variable

torch.manual_seed(0)

# Create a small LSTM encoder
class SmallLSTM(nn.Module):
    def __init__(self, input_size, hidden_size, num_layers=1):
        super(SmallLSTM, self).__init__()
        self.lstm = nn.LSTM(input_size, hidden_size, num_layers, batch_first=True)
    
    def forward(self, packed_input, hidden=None):
        output, hidden = self.lstm(packed_input, hidden)
        return output, hidden

# Parameters
input_size = 4
hidden_size = 5
num_layers = 1

enc_lstm = SmallLSTM(input_size, hidden_size, num_layers)

# Create input data
# name_inp_var: (total_names, max_name_len, input_size)
# We'll create 6 total names across 2 columns
batch_size = 2
max_name_len = 3
total_names = 6

name_inp_var = torch.randn(total_names, max_name_len, input_size)

# name_len: lengths of each name (6 values)
name_len = np.array([3, 2, 3, 1, 2, 3], dtype=np.int64)

# col_len: number of names per column (2 columns)
# First column has 3 names, second column has 3 names
col_len = np.array([3, 3], dtype=np.int64)

ARGS = {
    'name_inp_var': name_inp_var,
    'name_len': name_len,
    'col_len': col_len,
    'enc_lstm': enc_lstm
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
col_name_encode Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Baidi96/text2sql/sqlnet/model/modules/net_utils.py
code served (permissive licence) · get_code("0d171b3704b19708")
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
import numpy as np
from torch.autograd import Variable

torch.manual_seed(0)

# Build a minimal LSTM module that matches what run_lstm expects
class SimpleEncoder(nn.Module):
    def __init__(self, input_size, hidden_size):
        super(SimpleEncoder, self).__init__()
        self.lstm = nn.LSTM(input_size, hidden_size, batch_first=True)
    
    def forward(self, packed_input, hidden=None):
        output, hidden_state = self.lstm(packed_input, hidden)
        return output, hidden_state

# Parameters
input_size = 4
hidden_size = 5
embedding_dim = 6

# Create the encoder LSTM
enc_lstm = SimpleEncoder(input_size, hidden_size)

# Create name_inp_var: (total_names, seq_len, input_size)
# We'll have 5 total names across all columns
name_inp_var = torch.randn(5, 3, input_size)

# Create name_len: lengths of each name sequence
name_len = np.array([3, 2, 3, 1, 2], dtype=np.int64)

# Create col_len: number of names per column
# Column 0 has 2 names, Column 1 has 3 names
col_len = np.array([2, 3], dtype=np.int64)

ARGS = {
    'name_inp_var': name_inp_var,
    'name_len': name_len,
    'col_len': col_len,
    'enc_lstm': enc_lstm
}
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
load_data Ran
ran; our contract draft was wrong, not the code
Baidi96/text2sql/sqlnet/utils.py
code served (permissive licence) · get_code("d6fe93cd33bdc370")
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 json
import tempfile
import os

# Create temporary files with test data
temp_dir = tempfile.mkdtemp()

# Create SQL data file
sql_file = os.path.join(temp_dir, "sql_data.jsonl")
with open(sql_file, 'w') as f:
    for i in range(5):
        sql_entry = {
            "id": i,
            "table_id": f"table_{i % 2}",
            "query": f"SELECT * FROM table_{i % 2}"
        }
        f.write(json.dumps(sql_entry) + '\n')

# Create table data file
table_file = os.path.join(temp_dir, "table_data.jsonl")
with open(table_file, 'w') as f:
    for i in range(2):
        table_entry = {
            "id": f"table_{i}",
            "name": f"table_{i}",
            "columns": ["col1", "col2", "col3"]
        }
        f.write(json.dumps(table_entry) + '\n')

ARGS = {
    "sql_paths": [sql_file],
    "table_paths": [table_file],
    "use_small": False
}
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
load_data Ran
ran; our contract draft was wrong, not the code
CX000/sqlnet_inference_py36/sqlnet/utils.py
code served (permissive licence) · get_code("76114d4f76721e22")
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 json
import tempfile
import os

# Create temporary SQL file with test data
sql_file = tempfile.NamedTemporaryFile(mode='w', suffix='.jsonl', delete=False)
sql_data_lines = [
    json.dumps({"table_id": "table1", "query": "SELECT * FROM users"}),
    json.dumps({"table_id": "table2", "query": "SELECT * FROM orders"}),
]
for line in sql_data_lines:
    sql_file.write(line + '\n')
sql_file.close()
sql_path = sql_file.name

# Create temporary table file with test data
table_file = tempfile.NamedTemporaryFile(mode='w', suffix='.jsonl', delete=False)
table_data_lines = [
    json.dumps({"id": "table1", "name": "users", "columns": ["id", "name"]}),
    json.dumps({"id": "table2", "name": "orders", "columns": ["order_id", "user_id"]}),
]
for line in table_data_lines:
    table_file.write(line + '\n')
table_file.close()
table_path = table_file.name

ARGS = {
    "sql_paths": sql_path,
    "table_paths": table_path,
    "use_small": False,
}
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
load_data Ran
ran; our contract draft was wrong, not the code
PriyankaDatar/NLP_SQL_Project/sqlnet/utils.py
code served (permissive licence) · get_code("9c57d0d5b33162aa")
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 json
import tempfile
import os

# Create temporary files with JSON data
temp_dir = tempfile.mkdtemp()

# Create SQL data file
sql_file = os.path.join(temp_dir, 'sql_data.jsonl')
with open(sql_file, 'w') as f:
    sql_records = [
        {'table_id': 'table_1', 'query': 'SELECT * FROM users'},
        {'table_id': 'table_2', 'query': 'SELECT * FROM orders'},
        {'table_id': 'table_1', 'query': 'SELECT name FROM users WHERE id > 5'},
    ]
    for record in sql_records:
        f.write(json.dumps(record) + '\n')

# Create table data file
table_file = os.path.join(temp_dir, 'table_data.jsonl')
with open(table_file, 'w') as f:
    table_records = [
        {'id': 'table_1', 'name': 'users', 'columns': ['id', 'name']},
        {'id': 'table_2', 'name': 'orders', 'columns': ['order_id', 'user_id']},
    ]
    for record in table_records:
        f.write(json.dumps(record) + '\n')

ARGS = {
    'sql_paths': sql_file,
    'table_paths': table_file,
    'use_small': False,
}
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
load_dataset_dummy Ran
ran; our contract draft was wrong, not the code
CX000/sqlnet_inference_py36/sqlnet/utils.py
code served (permissive licence) · get_code("be979946e6e72c09")
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 json
import os
import tempfile

# Create temporary directory and mock data files
tmpdir = tempfile.mkdtemp()
mock_dir = os.path.join(tmpdir, 'mock')
os.makedirs(mock_dir, exist_ok=True)

# Create dummy SQL data file
sql_data = [
    {"table_id": "table_1", "query": "SELECT * FROM users", "sql": "select * from users"},
    {"table_id": "table_2", "query": "SELECT name FROM products", "sql": "select name from products"},
]
sql_file = os.path.join(mock_dir, 'dummy_tok.jsonl')
with open(sql_file, 'w') as f:
    for item in sql_data:
        f.write(json.dumps(item) + '\n')

# Create dummy table data file
table_data = [
    {"id": "table_1", "name": "users", "columns": ["id", "name"]},
    {"id": "table_2", "name": "products", "columns": ["id", "name", "price"]},
]
table_file = os.path.join(mock_dir, 'dummy_tok.tables.jsonl')
with open(table_file, 'w') as f:
    for item in table_data:
        f.write(json.dumps(item) + '\n')

# Also create the teststr variant file for completeness
sql_file_teststr = os.path.join(mock_dir, 'dummy_tok_test.jsonl')
with open(sql_file_teststr, 'w') as f:
    for item in sql_data:
        f.write(json.dumps(item) + '\n')

# Change to temp directory so relative paths work
original_dir = os.getcwd()
os.chdir(tmpdir)

ARGS = {
    'dataset_id': 0,
    'use_small': False,
    'teststr': ''
}
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
run_lstm Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
PriyankaDatar/NLP_project_BIDAF/sqlnet/model/modules/net_utils.py
code served (permissive licence) · get_code("a05f53e403297257")
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) 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
import numpy as np

torch.manual_seed(0)

# Create a small LSTM module
lstm = nn.LSTM(input_size=4, hidden_size=3, num_layers=1, batch_first=True)

# Create input tensor: batch_size=2, seq_len=5, input_size=4
inp = torch.randn(2, 5, 4)

# Create input lengths array: different lengths for each sequence in batch
inp_len = np.array([5, 3], dtype=np.int64)

# Create hidden state: (num_layers, batch_size, hidden_size)
h_0 = torch.randn(1, 2, 3)
c_0 = torch.randn(1, 2, 3)
hidden = (h_0, c_0)

ARGS = {
    'lstm': lstm,
    'inp': inp,
    'inp_len': inp_len,
    'hidden': hidden
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
run_lstm Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Baidi96/text2sql/sqlnet/model/modules/net_utils.py
code served (permissive licence) · get_code("9a2f05c9b4e0a6ed")
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
import numpy as np

torch.manual_seed(0)

# Create a small LSTM module
lstm = nn.LSTM(input_size=4, hidden_size=3, num_layers=1, batch_first=True)

# Create input tensor: batch_size=2, max_seq_len=5, input_size=4
inp = torch.randn(2, 5, 4)

# Create input lengths for each sequence in batch
inp_len = np.array([5, 3])

# Create hidden state: (num_layers=1, batch_size=2, hidden_size=3)
h_0 = torch.randn(1, 2, 3)
c_0 = torch.randn(1, 2, 3)
hidden = (h_0, c_0)

ARGS = {
    'lstm': lstm,
    'inp': inp,
    'inp_len': inp_len,
    'hidden': hidden
}
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
variable Ran
ran, honoured the contract we drafted
CX000/sqlnet_inference_py36/ewc.py
pointer only (licence: BSD-3-Clause); the stored copy get_code serves, from oakdata/benchmark, is pointer only (licence: NONE) · get_code("d95385178f502fa2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (BSD-3-Clause), but get_code serves the stored copy, first harvested from oakdata/benchmark, 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.

check_acc Not yet run
unverified (harvested, no recorded run)
Baidi96/text2sql/check_acc/check_acc.py
code served (permissive licence) · get_code("6aaa6080fafbce8a")
How it ran

no invocation recorded.

col_name_encode Not yet run
unverified (harvested, no recorded run)
CX000/sqlnet_inference_py36/sqlnet/model/modules/net_utils.py
code served (permissive licence) · get_code("477cb473548b1048")
How it ran

no invocation recorded.

ewc_train Not yet run
unverified (harvested, no recorded run)
CX000/sqlnet_inference_py36/ewc.py
code served (permissive licence) · get_code("051d5e10e30db060")
How it ran

no invocation recorded.

load_dataset Not yet run
unverified (harvested, no recorded run)
Baidi96/text2sql/sqlnet/utils.py
code served (permissive licence) · get_code("e02e4d6dd9d1c2ce")
How it ran

no invocation recorded.

load_dataset Not yet run
unverified (harvested, no recorded run)
CX000/sqlnet_inference_py36/sqlnet/utils.py
code served (permissive licence) · get_code("43888b7a46c23bcf")
How it ran

no invocation recorded.

load_dataset Not yet run
unverified (harvested, no recorded run)
PriyankaDatar/NLP_SQL_Project/sqlnet/utils.py
code served (permissive licence) · get_code("7172d1aead4de6bd")
How it ran

no invocation recorded.

normal_train Not yet run
unverified (harvested, no recorded run)
CX000/sqlnet_inference_py36/ewc.py
code served (permissive licence) · get_code("61d0dcf14b7919a6")
How it ran

no invocation recorded.

run_GRU Not yet run
unverified (harvested, no recorded run)
PriyankaDatar/NLP_Project_Modfications/sqlnet/model/modules/net_utils.py
code served (permissive licence) · get_code("6dc375fd6ccbbc87")
How it ran

no invocation recorded.

Also present in other papers' code

Of 19 samples lifted from this paper's linked repositories, 19 carry a function identity and 0 do not. Of the 16 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 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.

FunctionAlso present inEarliest-datedOthers named
variable 4 other papers
4 later
2004.14366 (2020-04)
Elastic weight consolidation for better bias inoculation
2103.12827 (2021-03)
Fisher Task Distance and Its Application in Neural Architecture Search
2108.11005 (2021-08)
Wanderlust: Online Continual Object Detection in the Real World
and 1 more not named here

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

Synthesizing SQL queries from natural language is a long-standing open problem and has been attracting considerable interest recently. Toward solving the problem, the de facto approach is to employ a sequence-to-sequence-style model. Such an approach will necessarily require the SQL queries to be serialized. Since the same SQL query may have multiple equivalent serializations, training a sequence-to-sequence-style model is sensitive to the choice from one of them. This phenomenon is documented as the "order-matters" problem. Existing state-of-the-art approaches rely on reinforcement learning to reward the decoder when it generates any of the equivalent serializations. However, we observe that the improvement from reinforcement learning is limited. In this paper, we propose a novel approach, i.e., SQLNet, to fundamentally solve this problem by avoiding the sequence-to-sequence structure when the order does not matter. In particular, we employ a sketch-based approach where the sketch contains a dependency graph so that one prediction can be done by taking into consideration only the previous predictions that it depends on. In addition, we propose a sequence-to-set model as well as the column attention mechanism to synthesize the query based on the sketch. By combining all these novel techniques, we show that SQLNet can outperform the prior art by 9% to 13% on the WikiSQL task.

For agents

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

get_harvested_code_for_paper("1711.04436")
get_code_for_paper("1711.04436")
have("1711.04436")

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

curl https://syntology.ai/api/ran/1711.04436.json

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

[![Syntology run record](https://syntology.ai/api/ran/1711.04436.svg)](https://syntology.ai/paper/1711.04436)

Connect an agent — have() is free.