SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1803.08475 · 2018

Attention, Learn to Solve Routing Problems!

arXiv · PDF · Open in the Atlas

Code that ran

27 of 40 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; 3 honoured the contract we drafted, 0 violated it and 13 have no contract verdict; 11 failures of our instrument; 13 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/1803.08475.json · badge: /api/ran/1803.08475.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
wouterkool/attention-tsp canonical (classifier label)
classifier basis not recorded on the link
3 ran · 2 no recorded run
tonydavis629/haucs community (classifier label)
classifier basis not recorded on the link
7 ran · 4 no recorded run
theresearchai/vehicle_routing_rl_2 community (classifier label)
classifier basis: plan_row
6 ran · 5 no recorded run
Rintarooo/VRP_DRL_MHA community (classifier label)
classifier basis not recorded on the link
4 ran · 1 no recorded run
raphaelavalos/attention_tsp_graph_net reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
wpwei/attention_to_route community (classifier label)
classifier basis not recorded on the link
2 ran · 1 no recorded run
eharmonicminorperfect5thbelow/pytorch-attention-model-tsp reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
FunctionStatusWhere it lives
CRC8_table_lookup Ran
ran on a synthesized input
tonydavis629/haucs/haucs/mission_planner/splashdrone.py
code served (permissive licence) · get_code("f25ac4888d367dac")
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
import numpy as np

# The function CRC8_table_lookup takes a buffer (sequence of values) and an offset (integer)
# It iterates through the buffer starting from the offset position and computes a CRC8 checksum

# Create a simple buffer - can be a list, array, or bytes
buffer = [0x42, 0x73, 0x15, 0x88, 0xA3]

# Create an offset - must be a valid index such that buffer[i + offset] doesn't exceed bounds
# Since the loop goes from 0 to len(buffer)-1, we need offset + (len(buffer)-1) < len(buffer)
# So offset should be 0 or negative (but that doesn't make sense for this use case)
# Actually, looking more carefully: for i in range(len(buffer)), it accesses buffer[i + offset]
# So if buffer has length 5 and offset is 0, it accesses indices 0,1,2,3,4 - valid
offset = 0

ARGS = {
    "buffer": buffer,
    "offset": offset
}
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 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
TSP Ran
ran on a synthesized input
wpwei/attention_to_route/problems.py
code served (permissive licence) · get_code("64903ec0e6fee27f")
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 2, dtype family float, shape (2, 1) 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)

# TSP expects:
# 1. graphs: a tensor that will be gathered from using permutations as indices
#    - Shape should be (batch_size, num_nodes, 2) for 2D coordinates
# 2. permutations: a tensor of shape (batch_size, num_nodes) containing valid tours
#    - Must be a permutation of [0, 1, ..., num_nodes-1] for each batch element

batch_size = 2
num_nodes = 4

# Create graphs as (batch_size, num_nodes, 2) - coordinates of nodes
# This represents 2D point coordinates for TSP
graphs = torch.tensor([
    [[0.0, 0.0], [1.0, 0.0], [1.0, 1.0], [0.0, 1.0]],
    [[0.5, 0.5], [1.5, 0.5], [1.5, 1.5], [0.5, 1.5]]
], dtype=torch.float32)

# Create valid permutations (tours) - each row is a permutation of 0..num_nodes-1
permutations = torch.tensor([
    [0, 1, 2, 3],
    [0, 2, 1, 3]
], dtype=torch.long)

ARGS = {
    "graphs": graphs,
    "permutations": permutations
}
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 Tensor, ndim 2, dtype family float, shape (2, 1) — shape (2, 1) 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
check_extension Ran
ran; our contract draft was wrong, not the code
theresearchai/vehicle_routing_rl_2/utils/data_utils.py
code served (permissive licence) · get_code("01df1926cf07ac89")
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 os

def check_extension(filename):
    if os.path.splitext(filename)[1] != '.pkl':
        return filename + '.pkl'
    return filename

ARGS = {
    'filename': 'myfile.txt'
}
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 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
coord2arr Ran
ran on a synthesized input
tonydavis629/haucs/haucs/utils/utils.py
code served (permissive licence) · get_code("1004553fb8c91911")
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 numpy as np

# Create sample latitude/longitude coordinates
# Using realistic coordinate ranges: latitudes roughly -90 to 90, longitudes -180 to 180
coords = np.array([
    [40.7128, -74.0060],   # New York City
    [34.0522, -118.2437],  # Los Angeles
    [41.8781, -87.6298],   # Chicago
    [29.7604, -95.3698],   # Houston
])

ARGS = {
    "coords": coords
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
data_from_txt Ran
ran on a synthesized input
Rintarooo/VRP_DRL_MHA/PyTorch/data.py
code served (permissive licence) · get_code("8daaa059e9c3cb2d")
How it ran

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

import torch
import numpy as np
import tempfile
import os

torch.manual_seed(0)

# Create a temporary text file with valid VRP data format
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_path = temp_file.name

# Write valid VRP data in the expected format
vrp_data = """NAME : test
COMMENT : test problem
TYPE : CVRP
DIMENSION : 4
EDGE_WEIGHT_TYPE : EUC_2D
CAPACITY : 100
NODE_COORD_SECTION
1 50.0 50.0
2 100.0 100.0
3 150.0 150.0
4 200.0 200.0
DEMAND_SECTION
1 0
2 10.0
3 20.0
4 30.0
DEPOT_SECTION
1
EOF
"""

temp_file.write(vrp_data)
temp_file.close()

ARGS = {
    'path': temp_path
}
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
evaluate Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
eharmonicminorperfect5thbelow/pytorch-attention-model-tsp/tsp.py
code served (permissive licence) · get_code("92bf6e766294e2b3")
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 float, 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 torch

torch.manual_seed(0)

# instances: a tensor of shape (num_instances, size, feature_dim)
# The function indexes into instances[i][solutions[i][j]], so:
# - instances needs at least 2 dimensions for indexing [i][j]
# - instances[i] should be indexable by solutions[i][j]
# - Each point should be a vector that supports arithmetic and .sum()

# solutions: a tensor or list of shape (num_instances, size) containing indices
# The function uses solutions[i][j] to index into instances[i]
# Each element should be an integer in range [0, size)

num_instances = 2
size = 4
feature_dim = 3

instances = torch.randn(num_instances, size, feature_dim)
solutions = torch.tensor([[0, 1, 2, 3], [3, 2, 1, 0]])

ARGS = {
    "instances": instances,
    "solutions": solutions
}
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 float, 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
generate_data Ran
ran on a synthesized input
Rintarooo/VRP_DRL_MHA/PyTorch/data.py
code served (permissive licence) · get_code("a58fa4a05eb041d8")
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)

ARGS = {
    'device': 'cpu',
    'n_samples': 2,
    'n_customer': 20,
    'seed': 42
}
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
generate_instances Ran
ran, honoured the contract we drafted
eharmonicminorperfect5thbelow/pytorch-attention-model-tsp/tsp.py
code served (permissive licence) · get_code("7c1af2ae74b68db5")
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 3, dtype family float, shape (2, 4, 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 torch

torch.manual_seed(0)

ARGS = {
    "size": 4,
    "num": 2,
}
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 3, dtype family float, shape (2, 4, 2) — shape (2, 4, 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
generate_op_data Ran
ran; our contract draft was wrong, not the code
wouterkool/attention-tsp/generate_data.py
code served (permissive licence) · get_code("82eaf1d494d4fc16")
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 list. 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 numpy as np

ARGS = {
    'dataset_size': 2,
    'op_size': 20,
    'prize_type': 'const'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list 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
generate_tsp_data Ran
ran, honoured the contract we drafted
wouterkool/attention-tsp/generate_data.py
code served (permissive licence) · get_code("50522f260eb670be")
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 list, ndim 3, 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

ARGS = {
    "dataset_size": 5,
    "tsp_size": 8,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 3, 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
generate_vrp_data Ran
ran; our contract draft was wrong, not the code
wouterkool/attention-tsp/generate_data.py
code served (permissive licence) · get_code("3a929eede11222c2")
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 list. 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 numpy as np

ARGS = {
    "dataset_size": 2,
    "vrp_size": 10,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_device Ran
ran; our contract draft was wrong, not the code
raphaelavalos/attention_tsp_graph_net/train_single_gpu.py
pointer only (licence: GPL-3.0) · get_code("08cb493b027671f6")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, raphaelavalos/attention_tsp_graph_net, is not in the permissive set Syntology serves inline. 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.

get_dist Ran
ran on a synthesized input
Rintarooo/VRP_DRL_MHA/PyTorch/dist_matrix.py
code served (permissive licence) · get_code("25f758eab8954cab")
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
import math
import numpy as np

torch.manual_seed(0)

# Test with torch.Tensor
n1_tensor = torch.tensor([0.0, 0.0, 3.0, 4.0])
n2_tensor = torch.tensor([1.0, 1.0, 4.0, 5.0])

# Test with list
n1_list = [0.0, 0.0, 3.0, 4.0]
n2_list = [1.0, 1.0, 4.0, 5.0]

# Test with numpy array
n1_numpy = np.array([0.0, 0.0, 3.0, 4.0])
n2_numpy = np.array([1.0, 1.0, 4.0, 5.0])

ARGS = {
    "n1": n1_tensor,
    "n2": n2_tensor,
}
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 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
get_dist_matrix Ran
ran on a synthesized input
Rintarooo/VRP_DRL_MHA/PyTorch/dist_matrix.py
code served (permissive licence) · get_code("f8bd592f10d20c70")
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 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 torch
import numpy as np
import math

torch.manual_seed(0)

# Create sample points - need 4 coordinates per point (x1, y1, x2, y2)
# Test with different point representations

# Points as list of lists
points_list = [
    [0, 0, 1, 1],
    [1, 1, 2, 2],
    [3, 3, 4, 4]
]

# Points as numpy arrays
points_numpy = [
    np.array([0.0, 0.0, 1.0, 1.0]),
    np.array([1.0, 1.0, 2.0, 2.0]),
    np.array([3.0, 3.0, 4.0, 4.0])
]

# Points as torch tensors
points_torch = [
    torch.tensor([0.0, 0.0, 1.0, 1.0]),
    torch.tensor([1.0, 1.0, 2.0, 2.0]),
    torch.tensor([3.0, 3.0, 4.0, 4.0])
]

ARGS = {
    "points": points_list,
    "digit": 2
}
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 2, 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_rollout_np Ran
ran, honoured the contract we drafted
raphaelavalos/attention_tsp_graph_net/train_single_gpu.py
pointer only (licence: GPL-3.0) · get_code("ddda4a28f60dd187")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, raphaelavalos/attention_tsp_graph_net, is not in the permissive set Syntology serves inline. 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.

get_validation_np Ran
ran; our contract draft was wrong, not the code
raphaelavalos/attention_tsp_graph_net/train_single_gpu.py
pointer only (licence: GPL-3.0) · get_code("fd794d93df5fe1c0")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, raphaelavalos/attention_tsp_graph_net, is not in the permissive set Syntology serves inline. 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.

load_data_model Ran
ran on a synthesized input
tonydavis629/haucs/haucs/solvers/googleORsolver.py
code served (permissive licence) · get_code("bfdb082b8b60d5ae")
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 pickle
import tempfile
import os

# Create a temporary pickle file with sample data
temp_dir = tempfile.gettempdir()
temp_filename = os.path.join(temp_dir, "test_haucs_data.pkl")

# Create sample data structure
sample_data = {
    "features": [1, 2, 3, 4, 5],
    "labels": [0, 1, 0, 1, 0],
    "metadata": {"source": "HAUCS", "version": 1}
}

# Write the sample data to a pickle file
with open(temp_filename, 'wb') as f:
    pickle.dump(sample_data, f)

# Define ARGS mapping all parameters
ARGS = {
    "filename": temp_filename
}
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 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
load_dataset Ran
ran; our contract draft was wrong, not the code
theresearchai/vehicle_routing_rl_2/utils/data_utils.py
code served (permissive licence) · get_code("69062811f34d8ff6")
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 os
import pickle
import tempfile

# Create a temporary pickle file for testing
temp_dir = tempfile.mkdtemp()
test_data = {"sample": "data", "values": [1, 2, 3]}
test_file_path = os.path.join(temp_dir, "test_dataset.pkl")

with open(test_file_path, 'wb') as f:
    pickle.dump(test_data, f)

# Remove the .pkl extension for the argument to test the check_extension logic
test_filename = os.path.join(temp_dir, "test_dataset")

ARGS = {
    "filename": test_filename
}
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 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
load_dataset Ran
ran on a synthesized input
wpwei/attention_to_route/utils.py
code served (permissive licence) · get_code("c43c60dd79880c30")
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, 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 os
import pickle
import numpy as np
import tempfile

# Create a temporary directory and a test pickle file
temp_dir = tempfile.mkdtemp()
test_data = [[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]]
test_filename = os.path.join(temp_dir, 'test_dataset')

# Write the test data to a pickle file
with open(test_filename + '.pkl', 'wb') as f:
    pickle.dump(test_data, f)

ARGS = {
    'filename': test_filename
}
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, 3) — shape (2, 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
load_files Ran
ran on a synthesized input
tonydavis629/haucs/haucs/mission_planner/splashdrone.py
code served (permissive licence) · get_code("222a077ac0dc342e")
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 3, 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 tempfile
import os

# Create a temporary directory and sample route files
temp_dir = tempfile.mkdtemp()

# Create sample route files
route_name = "route_"
for i in range(3):
    filepath = os.path.join(temp_dir, route_name + str(i) + '.txt')
    with open(filepath, 'w') as f:
        f.write("1.0,2.0,3.0\n")
        f.write("4.0,5.0,6.0\n")
        f.write("7.0,8.0,9.0\n")

ARGS = {
    'save_dir': temp_dir + '/',
    'route_name': route_name
}
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 list, ndim 3, 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
load_pts Ran
ran on a synthesized input
tonydavis629/haucs/haucs/mission_planner/splashdrone.py
code served (permissive licence) · get_code("7432af444f223046")
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 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 tempfile
import os

# Create a temporary file with point data
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "points.txt")

with open(temp_file, 'w') as f:
    f.write("1.0,2.0,3.0\n")
    f.write("4.5,5.5,6.5\n")
    f.write("7.2,8.3,9.1\n")

ARGS = {
    "filename": temp_file
}
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 list, ndim 2, 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
mask_bool2long Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
theresearchai/vehicle_routing_rl_2/utils/boolmask.py
code served (permissive licence) · get_code("454003b7383dce9f")
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 2, dtype family int, shape (2, 1) 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.functional as F

torch.manual_seed(0)

# Create a small uint8 tensor for mask input
# The function expects a uint8 tensor with shape (..., sequence_length)
# Let's create a tensor with batch_size=2 and sequence_length=4
mask = torch.randint(0, 2, (2, 4), dtype=torch.uint8)

ARGS = {
    "mask": mask
}
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 2, dtype family int, shape (2, 1) — shape (2, 1) 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
mask_long2bool Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
theresearchai/vehicle_routing_rl_2/utils/boolmask.py
code served (permissive licence) · get_code("10638ccfaa13fb59")
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 2, dtype family float, shape (2, 256) 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)

# mask_long2bool takes a mask tensor of dtype int64 and optional n parameter
# The function converts int64 mask to boolean mask by bit manipulation

# Create a small int64 tensor for the mask
# Shape: (2, 4) - batch of 2, 4 int64 values
mask = torch.tensor([[1, 2, 4, 8], [16, 32, 64, 128]], dtype=torch.int64)

# n is optional - can be None or an integer specifying number of bits
n = None

ARGS = {
    "mask": mask,
    "n": n
}
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 2, dtype family float, shape (2, 256) — shape (2, 256) 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
mask_long_scatter Ran
ran; our contract draft was wrong, not the code
theresearchai/vehicle_routing_rl_2/utils/boolmask.py
code served (permissive licence) · get_code("1a3bb2f0ebdf9a4e")
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 3, dtype family int, 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

torch.manual_seed(0)

# Create a mask tensor with batch dimensions and last dimension for bit packing
# mask shape: (2, 3, 4) - batch dims (2, 3), last dim 4 for bit positions
mask = torch.zeros((2, 3, 4), dtype=torch.long)

# Create values tensor with same shape as mask[:-1]
# values shape: (2, 3) - must match mask.size()[:-1]
# Values should be in range [0, 256) to work with the 64-bit packing logic
values = torch.tensor([[10, 50, 120], [200, 30, 100]], dtype=torch.long)

# check_unset parameter - boolean flag
check_unset = False

ARGS = {
    "mask": mask,
    "values": values,
    "check_unset": check_unset,
}
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 3, dtype family int, 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
move_to Ran
ran; our contract draft was wrong, not the code
theresearchai/vehicle_routing_rl_2/utils/functions.py
pointer only (licence: MIT); the stored copy get_code serves, from zhangzizhen/ml-dam, is pointer only (licence: NONE) · get_code("107d6628712f1a21")
How it ran

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

normalize Ran
ran on a synthesized input
tonydavis629/haucs/haucs/utils/utils.py
code served (permissive licence) · get_code("c33c1c373528994c")
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 (6,) 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

# The function normalize takes a single parameter 'x' which should be array-like
# It converts x to a numpy array, then normalizes it by subtracting the minimum
# and dividing by the peak-to-peak (range) value

# Create test data - a simple array with various values
test_array = np.array([1.0, 2.5, 3.0, 1.5, 4.2, 0.5])

ARGS = {
    'x': test_array
}
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 (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
processGPS Ran
ran on a synthesized input
tonydavis629/haucs/mobile_app/datalogger.py
code served (permissive licence) · get_code("3855f5b5d541ad04")
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 numpy as np

torch.manual_seed(0)

# Analyze the function:
# - processGPS takes one parameter: message
# - message is expected to be a list
# - The function iterates through message starting at index 1, pairing elements
# - It expects message to contain key-value pairs where keys are strings
# - Specifically, it expects 'DVAL' and 'TVAL' keys in the resulting dict
# - These are popped and concatenated to form the timestamp
# - The function returns (data dict, timestamp string)

# Construct a valid message list:
# Format: [anything, key1, value1, key2, value2, ..., 'DVAL', dval_str, 'TVAL', tval_str]
# The first element is skipped (i starts at 1)

message = [
    'IGNORED_FIRST_ELEMENT',
    'LAT', '40.7128',
    'LON', '-74.0060',
    'DVAL', '2024-01-15',
    'TVAL', '14:30:45'
]

ARGS = {
    'message': message
}
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, 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
backtrack Not yet run
unverified (harvested, no recorded run)
theresearchai/vehicle_routing_rl_2/utils/beam_search.py
code served (permissive licence) · get_code("9d5b8d4a73d01c19")
How it ran

no invocation recorded.

gen_results Not yet run
unverified (harvested, no recorded run)
tonydavis629/haucs/haucs/solvers/googleORsolver.py
code served (permissive licence) · get_code("82c1f4dc5f78132f")
How it ran

no invocation recorded.

get_beam_search_results Not yet run
unverified (harvested, no recorded run)
theresearchai/vehicle_routing_rl_2/utils/beam_search.py
code served (permissive licence) · get_code("be9c547a8e748da5")
How it ran

no invocation recorded.

get_options Not yet run
unverified (harvested, no recorded run)
theresearchai/vehicle_routing_rl_2/options.py
code served (permissive licence) · get_code("0f6f8675d8280c4f")
How it ran

no invocation recorded.

init_file Not yet run
unverified (harvested, no recorded run)
tonydavis629/haucs/mobile_app/datalogger.py
code served (permissive licence) · get_code("646e8f569fdfb4c8")
How it ran

no invocation recorded.

init_file Not yet run
unverified (harvested, no recorded run)
tonydavis629/haucs/basestation/pc_basestation/datalogger.py
code served (permissive licence) · get_code("63a2c7b484c8eb97")
How it ran

no invocation recorded.

lkh_solve_tsp Not yet run
unverified (harvested, no recorded run)
wpwei/attention_to_route/utils.py
code served (permissive licence) · get_code("a9dc1cf1047bd89e")
How it ran

no invocation recorded.

load_pkl Not yet run
unverified (harvested, no recorded run)
Rintarooo/VRP_DRL_MHA/PyTorch/config.py
code served (permissive licence) · get_code("b71b6ddb1c0ba51a")
How it ran

no invocation recorded.

load_problem Not yet run
unverified (harvested, no recorded run)
theresearchai/vehicle_routing_rl_2/utils/functions.py
code served (permissive licence) · get_code("5deb1f5b2269293b")
How it ran

no invocation recorded.

save_norm_coords Not yet run
unverified (harvested, no recorded run)
tonydavis629/haucs/haucs/utils/utils.py
code served (permissive licence) · get_code("cf9bc36f5580661b")
How it ran

no invocation recorded.

solve_all_gurobi Not yet run
unverified (harvested, no recorded run)
wouterkool/attention-tsp/problems/tsp/tsp_gurobi.py
code served (permissive licence) · get_code("c2eb1a9181034240")
How it ran

no invocation recorded.

solve_euclidian_tsp Not yet run
unverified (harvested, no recorded run)
wouterkool/attention-tsp/problems/tsp/tsp_gurobi.py
code served (permissive licence) · get_code("c2edba20f4705527")
How it ran

no invocation recorded.

torch_load_cpu Not yet run
unverified (harvested, no recorded run)
theresearchai/vehicle_routing_rl_2/utils/functions.py
pointer only (licence: MIT); the stored copy get_code serves, from zhangzizhen/ml-dam, is pointer only (licence: NONE) · get_code("0538f709044ed305")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from zhangzizhen/ml-dam, 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. No invocation recorded.

Also present in other papers' code

Of 40 samples lifted from this paper's linked repositories, 40 carry a function identity and 0 do not. Of the 39 distinct functions with an identity, 16 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 16 later) and 23 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
check_extension 7 other papers
7 later
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2305.19587 (2023-05)
Towards Omni-generalizable Neural Methods for Vehicle Routing Problems
2308.14104 (2023-08)
Towards Generalizable Neural Solvers for Vehicle Routing Problems via Ensemble with Transferrable Local Policy
and 4 more not named here
load_dataset 4 other papers
4 later
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2403.07041 (2024-03)
Ant Colony Sampling with GFlowNets for Combinatorial Optimization
2410.21066 (2024-10)
Learning to Handle Complex Constraints for Vehicle Routing Problems
and 1 more not named here
torch_load_cpu 4 other papers
4 later
2105.02741 (2021-05)
Meta-Learning-Based Deep Reinforcement Learning for Multiobjective Optimization Problems
2110.14074 (2021-10)
Fault-Tolerant Federated Reinforcement Learning with Theoretical Guarantee
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
and 1 more not named here
generate_op_data 3 other papers
3 later
2006.07054 (2020-06)
Learning the Travelling Salesperson Problem Requires Rethinking Generalization
2012.10638 (2020-12)
Multi-Decoder Attention Model with Embedding Glimpse for Solving Vehicle Routing Problems
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
generate_tsp_data 3 other papers
3 later
2006.07054 (2020-06)
Learning the Travelling Salesperson Problem Requires Rethinking Generalization
2012.10638 (2020-12)
Multi-Decoder Attention Model with Embedding Glimpse for Solving Vehicle Routing Problems
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
generate_vrp_data 3 other papers
3 later
2006.07054 (2020-06)
Learning the Travelling Salesperson Problem Requires Rethinking Generalization
2012.10638 (2020-12)
Multi-Decoder Attention Model with Embedding Glimpse for Solving Vehicle Routing Problems
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
move_to 3 other papers
3 later
2105.02741 (2021-05)
Meta-Learning-Based Deep Reinforcement Learning for Multiobjective Optimization Problems
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
backtrack 2 other papers
2 later
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
get_beam_search_results 2 other papers
2 later
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
mask_bool2long 2 other papers
2 later
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
mask_long2bool 2 other papers
2 later
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
mask_long_scatter 2 other papers
2 later
2205.13225 (2022-05)
DevFormer: A Symmetric Transformer for Context-Aware Device Placement
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
get_options 1 other paper
1 later
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
load_problem 1 other paper
1 later
2606.30680 (2026-06)
Locker-based Truck-Drone Routing with Integrated Considerations of Pickups, Deliveries, and No-Fly Zones
solve_all_gurobi 1 other paper
1 later
2006.07054 (2020-06)
Learning the Travelling Salesperson Problem Requires Rethinking Generalization
solve_euclidian_tsp 1 other paper
1 later
2006.07054 (2020-06)
Learning the Travelling Salesperson Problem Requires Rethinking Generalization

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 recently presented idea to learn heuristics for combinatorial optimization problems is promising as it can save costly development. However, to push this idea towards practical implementation, we need better models and better ways of training. We contribute in both directions: we propose a model based on attention layers with benefits over the Pointer Network and we show how to train this model using REINFORCE with a simple baseline based on a deterministic greedy rollout, which we find is more efficient than using a value function. We significantly improve over recent learned heuristics for the Travelling Salesman Problem (TSP), getting close to optimal results for problems up to 100 nodes. With the same hyperparameters, we learn strong heuristics for two variants of the Vehicle Routing Problem (VRP), the Orienteering Problem (OP) and (a stochastic variant of) the Prize Collecting TSP (PCTSP), outperforming a wide range of baselines and getting results close to highly optimized and specialized algorithms.

For agents

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

get_harvested_code_for_paper("1803.08475")
get_code_for_paper("1803.08475")
have("1803.08475")

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

curl https://syntology.ai/api/ran/1803.08475.json

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

[![Syntology run record](https://syntology.ai/api/ran/1803.08475.svg)](https://syntology.ai/paper/1803.08475)

Connect an agent — have() is free.