SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2510.20872 · NeurIPS · 2025

MOBO-OSD: Batch Multi-Objective Bayesian Optimization via Orthogonal Search Directions

Huong Ha, Jeffrey Chan, Lam Ngo, Hongyu Zhang

arXiv · PDF · Open in the Atlas

Code that ran

18 of 26 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, 1 violated it and 0 have no contract verdict; 14 failures of our instrument; 8 no recorded run), measured 2026-11-22; 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/2510.20872.json · badge: /api/ran/2510.20872.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
Alaleh/PDBO canonical (classifier label)
classifier basis: census
6 ran · 7 no recorded run
LamNgo1/mobo-osd canonical (classifier label)
classifier basis: census
6 ran · 1 no recorded run
belakaria/USeMO canonical (classifier label)
classifier basis: census
6 ran · 0 no recorded run
FunctionStatusWhere it lives
LCB Ran
ran; our contract draft was wrong, not the code
belakaria/USeMO/acquisitions.py
code served (permissive licence) · get_code("52d18a279f43e48d")
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 torch.nn as nn
import numpy as np

torch.manual_seed(0)

class SimpleGP(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear_mean = nn.Linear(4, 1)
        self.linear_std = nn.Linear(4, 1)
    
    def getPrediction(self, x):
        """Returns mean and std predictions for input x"""
        mean = self.linear_mean(x)
        std = torch.nn.functional.softplus(self.linear_std(x))
        return mean, std

# Create instances
gp = SimpleGP()
x = torch.randn(2, 4)
beta = 1.5

ARGS = {
    "x": x,
    "beta": beta,
    "GP": gp
}
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
Rosen Ran
ran; our contract draft was wrong, not the code
belakaria/USeMO/benchmarks.py
code served (permissive licence) · get_code("65d64cfeca11f079")
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 float64, ndim 0, 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

# x1 should be an array-like object that can be converted to a numpy array
# The function multiplies it by 4 and subtracts 2, so it should be numeric
# d is the dimension parameter that controls the loop range (range(d-1))

x1_value = np.array([0.5, 1.2, 0.8, 1.5])  # Array of 4 elements
d_value = 4  # Dimension parameter, must be at least 2 for the loop to execute

ARGS = {
    "x1": x1_value,
    "d": d_value
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
Sphere Ran
ran; our contract draft was wrong, not the code
belakaria/USeMO/benchmarks.py
code served (permissive licence) · get_code("fd5c6a9c2884ffe7")
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 float64, ndim 0, 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

# Analyze the function:
# - x1: passed to np.asarray(), so needs to be array-like
# - d: used in range(d), so needs to be an integer
# - The function accesses x[i] for i in range(d), so x1 must have at least d elements

# Create simple test values
x1_value = [1.0, 2.0, 3.0, 4.0, 5.0]  # list of 5 floats
d_value = 4  # dimension parameter, should be <= len(x1)

ARGS = {
    "x1": x1_value,
    "d": d_value
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
UCB Ran
ran; our contract draft was wrong, not the code
belakaria/USeMO/acquisitions.py
code served (permissive licence) · get_code("1acd60ceda2ff72b")
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, 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

torch.manual_seed(0)

# Create a minimal GP class that implements getPrediction
class MinimalGP(nn.Module):
    def __init__(self):
        super().__init__()
        self.register_buffer('mean_offset', torch.randn(4))
        self.register_buffer('std_scale', torch.abs(torch.randn(4)) + 0.1)
    
    def getPrediction(self, x):
        # x shape: (batch, 4)
        batch_size = x.shape[0]
        mean = x + self.mean_offset.unsqueeze(0)
        std = self.std_scale.unsqueeze(0).expand(batch_size, -1)
        return mean, std

# Construct arguments
x = torch.randn(2, 4)  # batch of 2, dimension 4
beta = 1.96  # typical UCB exploration parameter
GP = MinimalGP()

ARGS = {
    'x': x,
    'beta': beta,
    'GP': GP
}
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, 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
batch_distance_from_m_point_to_n_line Ran
ran, honoured the contract we drafted
LamNgo1/mobo-osd/mobo_osd/util.py
code served (permissive licence) · get_code("ff8f15536396f1a4")
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 numpy as np

# Construct small test data for the function
# __f: m points in 3D space -> shape (m, 3)
# __point_on_line: n points on lines -> shape (n, 3)
# __line_normal: n normal vectors for lines -> shape (n, 3)

m = 2  # number of points
n = 3  # number of lines
d = 3  # dimension

# Create m points
f = np.random.randn(m, d).astype(np.float32)

# Create n points on lines
point_on_line = np.random.randn(n, d).astype(np.float32)

# Create n line normal vectors
line_normal = np.random.randn(n, d).astype(np.float32)

ARGS = {
    "__f": f,
    "__point_on_line": point_on_line,
    "__line_normal": line_normal,
}
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
clone Ran
ran; our contract draft was wrong, not the code
Alaleh/PDBO/mobo/PDBO/single_gp.py
pointer only (licence: NONE) · get_code("b9b7e6dece0aba55")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

compute_ref_point Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
LamNgo1/mobo-osd/mobo_osd/util.py
code served (permissive licence) · get_code("52e449a6b76c0ce5")
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 (3,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create a small numpy array with shape (2, 3) representing 2 samples with 3 objectives
_y = np.array([[1.0, 2.0, 3.0], 
               [4.0, 5.0, 6.0]])

_minimize = True

ARGS = {
    "_y": _y,
    "_minimize": _minimize
}
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 (3,) — shape (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
distance_to_line Ran
ran, honoured the contract we drafted
LamNgo1/mobo-osd/mobo_osd/helper.py
code served (permissive licence) · get_code("8da3c38f4a3f224b")
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 (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

# Construct arguments for distance_to_line
# The function computes distance from points to a line defined by a point and normal vector

# _f: points from which to measure distance (2D array, shape: [n_points, n_dims])
_f = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0]
])

# _point_on_line: a point on the line (1D array, shape: [n_dims])
_point_on_line = np.array([0.0, 0.0, 0.0, 0.0])

# _line_normal: direction/normal vector of the line (1D array, shape: [n_dims])
_line_normal = np.array([1.0, 0.0, 0.0, 0.0])

ARGS = {
    "_f": _f,
    "_point_on_line": _point_on_line,
    "_line_normal": _line_normal
}
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 (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_acquisation_function Ran
ran; our contract draft was wrong, not the code
belakaria/USeMO/acquisitions.py
code served (permissive licence) · get_code("660355b1283fe2a8")
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 torch
import numpy as np
from scipy.stats import norm


class MockGP:
    """Mock Gaussian Process with required methods and attributes."""
    
    def __init__(self):
        # Initialize with some sample y values for ei and pi functions
        self.yValues = np.array([1.0, 2.5, 1.8, 3.2])
        
        # Create a simple mock model with sample_y method
        self.model = MockGPModel()
    
    def getPrediction(self, x):
        """
        Returns mean and std predictions.
        x should be a 1D array or similar.
        """
        # Simple mock: return deterministic values based on input
        mean = float(np.sin(np.sum(x)) + 1.0)
        std = 0.5
        return mean, std


class MockGPModel:
    """Mock GP model with sample_y method for TS acquisition function."""
    
    def sample_y(self, x):
        """
        Sample from the GP posterior.
        x should be reshaped to (1, -1).
        Returns array of samples.
        """
        # Simple mock: return a single sample
        return np.array([2.1])


# Construct arguments for get_acquisation_function
ARGS = {
    "acquisation_name": "UCB"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, torch — 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_acquisition_args Ran
ran; our contract draft was wrong, not the code
Alaleh/PDBO/arguments.py
pointer only (licence: NONE) · get_code("e4c4abc2fdd0d89f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_benchmark_functions Ran
ran; our contract draft was wrong, not the code
belakaria/USeMO/benchmarks.py
code served (permissive licence) · get_code("8bd2d93e2a6807f0")
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
import math

# The function get_benchmark_functions takes one parameter: names
# It's a string that specifies which benchmark functions to return
# The function splits the string by commas and looks up functions in a dict

ARGS = {
    'names': 'branin,Currin,Powell'
}
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_general_args Ran
ran; our contract draft was wrong, not the code
Alaleh/PDBO/arguments.py
pointer only (licence: NONE) · get_code("089687090d326a23")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_kernels Ran
ran; our contract draft was wrong, not the code
Alaleh/PDBO/mobo/PDBO/DPP.py
pointer only (licence: NONE) · get_code("534eb6a255bab9a4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_nadir_ideal_points Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
LamNgo1/mobo-osd/mobo_osd/helper.py
code served (permissive licence) · get_code("01393f9f210edd97")
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

# Construct a small 2D array with shape (5, 3)
# representing 5 points in 3-objective space
_f = np.array([
    [1.0, 2.0, 3.0],
    [0.5, 1.5, 2.5],
    [2.0, 3.0, 1.0],
    [1.5, 2.5, 2.0],
    [0.8, 1.2, 1.8]
])

ARGS = {
    "_f": _f,
    "minimize": True
}
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_surroagte_args Ran
ran; our contract draft was wrong, not the code
Alaleh/PDBO/arguments.py
pointer only (licence: NONE) · get_code("8bf7022e018126b1")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_xy_based_on_hvc Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
LamNgo1/mobo-osd/mobo_osd/util.py
code served (permissive licence) · get_code("77c3b237a35c4577")
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

# Set random seed for reproducibility
np.random.seed(0)

# Create small sample data
# _x: decision variables (design space), shape (n_samples, n_dims)
# _y: objectives to minimize, shape (n_samples, n_objectives)
n_samples = 8
n_dims = 4
n_objectives = 3

_x = np.random.randn(n_samples, n_dims).astype(np.float64)
_y = np.random.randn(n_samples, n_objectives).astype(np.float64)

# Make some points dominated to test the hypervolume contribution logic
_y[0] = np.array([1.0, 1.0, 1.0])
_y[1] = np.array([2.0, 2.0, 2.0])
_y[2] = np.array([0.5, 0.5, 0.5])
_y[3] = np.array([1.5, 1.5, 1.5])

# Number of points to select
_n = 4

ARGS = {
    "_x": _x,
    "_y": _y,
    "_n": _n,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pymoo — 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
normalize Ran
ran, honoured the contract we drafted
Alaleh/PDBO/mobo/PDBO/HedgeAdaptiveAcq.py
pointer only (licence: NONE) · get_code("98b13eb19c035fc9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

to_unit_cube Ran
ran, violated the contract we drafted
LamNgo1/mobo-osd/mobo_osd/helper.py
code served (permissive licence) · get_code("2e9112577ee9230e")
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, 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 numpy as np

torch.manual_seed(0)

# Analyze the function:
# to_unit_cube(x, lb, ub) performs: (x - lb) / (ub - lb)
# Parameters:
# - x: value(s) to normalize (tensor or array)
# - lb: lower bound (tensor or array)
# - ub: upper bound (tensor or array)
# All should be broadcastable or same shape

# Create small example tensors
x = torch.tensor([[1.0, 2.0, 3.0, 4.0],
                   [2.0, 3.0, 4.0, 5.0]])  # shape (2, 4)

lb = torch.tensor([[0.0, 0.0, 0.0, 0.0],
                    [0.0, 0.0, 0.0, 0.0]])  # shape (2, 4)

ub = torch.tensor([[10.0, 10.0, 10.0, 10.0],
                    [10.0, 10.0, 10.0, 10.0]])  # shape (2, 4)

ARGS = {
    'x': x,
    'lb': lb,
    'ub': ub
}
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, 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
calculate_matrix_sums Not yet run
unverified (harvested, no recorded run)
Alaleh/PDBO/mobo/PDBO/DPP.py
pointer only (licence: NONE) · get_code("abb3e2a8028471f0")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

estimate_mean_and_std Not yet run
unverified (harvested, no recorded run)
LamNgo1/mobo-osd/mobo_osd/gp.py
code served (permissive licence) · get_code("76dbcb49c04d9a62")
How it ran

no invocation recorded.

get_acquisition Not yet run
unverified (harvested, no recorded run)
Alaleh/PDBO/mobo/factory.py
pointer only (licence: NONE) · get_code("877dd5166ff9a87e")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_acquisition_function Not yet run
unverified (harvested, no recorded run)
Alaleh/PDBO/mobo/PDBO/HedgeAdaptiveAcq.py
pointer only (licence: NONE) · get_code("75e922b6f4492672")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_result_dir Not yet run
unverified (harvested, no recorded run)
Alaleh/PDBO/utils.py
pointer only (licence: NONE) · get_code("e1475905997064b7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_solver Not yet run
unverified (harvested, no recorded run)
Alaleh/PDBO/mobo/factory.py
pointer only (licence: NONE) · get_code("545970a789de193c")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

get_surrogate_model Not yet run
unverified (harvested, no recorded run)
Alaleh/PDBO/mobo/factory.py
pointer only (licence: NONE) · get_code("4d9faa74579a7156")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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.

setup_logger Not yet run
unverified (harvested, no recorded run)
Alaleh/PDBO/utils.py
pointer only (licence: NONE) · get_code("f1fef78471496569")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Alaleh/PDBO 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 26 samples lifted from this paper's linked repositories, 26 carry a function identity and 0 do not. Of the 26 distinct functions with an identity, 16 are also present in other papers whose code we hold (16 with at least one paper dated earlier, 0 the same month, 2 later) and 10 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
get_result_dir 2 other papers
1 dated earlier, 1 later
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
2608.15073 (2026-08)
BOCoDe: Engineering-Centered Benchmarking for Bayesian Optimization
setup_logger 2 other papers
1 dated earlier, 1 later
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
2608.15073 (2026-08)
BOCoDe: Engineering-Centered Benchmarking for Bayesian Optimization
LCB 1 other paper
1 dated earlier
2204.05944 (2022-04)
Uncertainty-Aware Search Framework for Multi-Objective Bayesian Optimization
UCB 1 other paper
1 dated earlier
2204.05944 (2022-04)
Uncertainty-Aware Search Framework for Multi-Objective Bayesian Optimization
calculate_matrix_sums 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
clone 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_acquisation_function 1 other paper
1 dated earlier
2204.05944 (2022-04)
Uncertainty-Aware Search Framework for Multi-Objective Bayesian Optimization
get_acquisition 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_acquisition_args 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_acquisition_function 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_general_args 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_kernels 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_solver 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_surroagte_args 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
get_surrogate_model 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization
normalize 1 other paper
1 dated earlier
2406.08799 (2024-06)
Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization

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

Bayesian Optimization (BO) is a powerful tool for optimizing expensive blackbox objective functions. While extensive research has been conducted on the single-objective optimization problem, the multi-objective optimization problem remains challenging. In this paper, we propose MOBO-OSD, a multi-objective Bayesian Optimization algorithm designed to generate a diverse set of Pareto optimal solutions by solving multiple constrained optimization problems, referred to as MOBO-OSD subproblems, along orthogonal search directions (OSDs) defined with respect to an approximated convex hull of individual objective minima. By employing a well-distributed set of OSDs, MOBO-OSD ensures broad coverage of the objective space, enhancing both solution diversity and hypervolume performance. To further improve the density of the set of Pareto optimal candidate solutions without requiring an excessive number of subproblems, we leverage a Pareto Front Estimation technique to generate additional solutions in the neighborhood of existing solutions. Additionally, MOBO-OSD supports batch optimization, enabling parallel function evaluations to accelerate the optimization process when resources are available. Through extensive experiments and analysis on a variety of synthetic and real-world benchmark functions with two to six objectives, we demonstrate that MOBO-OSD consistently outperforms the state-of-the-art algorithms. Our code implementation can be found at https://github.com/LamNgo1/mobo-osd.

For agents

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

get_harvested_code_for_paper("2510.20872")
get_code_for_paper("2510.20872")
have("2510.20872")

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

curl https://syntology.ai/api/ran/2510.20872.json

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

[![Syntology run record](https://syntology.ai/api/ran/2510.20872.svg)](https://syntology.ai/paper/2510.20872)

Connect an agent — have() is free.