SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.07144 · 2026

BONSAI: Bayesian Optimization with Natural Simplicity and Interpretability

Samuel Daulton, Meta Eriksson, Meta Balandat, Meta Eytan, Bakshy Meta

arXiv · PDF · Open in the Atlas

Code that ran

2 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 failures of our instrument; 6 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/2602.07144.json · badge: /api/ran/2602.07144.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
cjlin1/libsvm canonical (classifier label)
classifier basis: census
2 ran · 6 no recorded run
FunctionStatusWhere it lives
genFields Ran
ran; our contract draft was wrong, not the code
cjlin1/libsvm/python/libsvm/svm.py
pointer only (licence: BSD-3-Clause); the stored copy get_code serves, from xmc-aalto/dismec, is pointer only (licence: NONE) · get_code("2065a445394e47c6")
How it ran

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

toPyModel Ran
ran; our contract draft was wrong, not the code
cjlin1/libsvm/python/libsvm/svm.py
code served (permissive licence) · get_code("35651866707f2fb5")
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 svm_model. 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 ctypes

# The function expects a ctypes POINTER(svm_model).
# We need to create a minimal svm_model structure and a valid pointer to it.

# First, define a minimal svm_model class that can be used with ctypes
class svm_model(ctypes.Structure):
    """Minimal SVM model structure for testing"""
    pass

# Add some basic fields that an svm_model might have
svm_model._fields_ = [
    ('nr_class', ctypes.c_int),
    ('l', ctypes.c_int),
    ('SV', ctypes.POINTER(ctypes.c_double)),
    ('sv_coef', ctypes.POINTER(ctypes.POINTER(ctypes.c_double))),
    ('rho', ctypes.POINTER(ctypes.c_double)),
]

# Create an instance of svm_model
model_instance = svm_model()
model_instance.nr_class = 2
model_instance.l = 10

# Create a pointer to the model instance
model_ptr = ctypes.POINTER(svm_model)(model_instance)

ARGS = {
    'model_ptr': model_ptr
}
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 svm_model — a svm_model 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
evaluations Not yet run
unverified (harvested, no recorded run)
cjlin1/libsvm/python/libsvm/commonutil.py
code served (permissive licence) · get_code("10d70ae1c81cb35c")
How it ran

no invocation recorded.

evaluations_scipy Not yet run
unverified (harvested, no recorded run)
cjlin1/libsvm/python/libsvm/commonutil.py
code served (permissive licence) · get_code("0303193917372aa0")
How it ran

no invocation recorded.

svm_load_model Not yet run
unverified (harvested, no recorded run)
cjlin1/libsvm/python/libsvm/svmutil.py
code served (permissive licence) · get_code("5510ad1299b39c39")
How it ran

no invocation recorded.

svm_predict Not yet run
unverified (harvested, no recorded run)
cjlin1/libsvm/python/libsvm/svmutil.py
code served (permissive licence) · get_code("563a569e011fe403")
How it ran

no invocation recorded.

svm_read_problem Not yet run
unverified (harvested, no recorded run)
cjlin1/libsvm/python/libsvm/commonutil.py
code served (permissive licence) · get_code("7bb322a885f10539")
How it ran

no invocation recorded.

svm_train Not yet run
unverified (harvested, no recorded run)
cjlin1/libsvm/python/libsvm/svmutil.py
code served (permissive licence) · get_code("a2c959c492da6559")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 0 later) and 5 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
genFields 5 other papers
5 dated earlier
1609.02521 (2016-09)
DiSMEC - Distributed Sparse Machines for Extreme Multi-label Classification
1907.12865 (2019-07)
Open Set Domain Adaptation for Image and Action Recognition
2004.09930 (2020-04)
Relabel the Noise: Joint Extraction of Entities and Relations via Cooperative Multiagents
and 2 more not named here
toPyModel 5 other papers
5 dated earlier
1609.02521 (2016-09)
DiSMEC - Distributed Sparse Machines for Extreme Multi-label Classification
1907.12865 (2019-07)
Open Set Domain Adaptation for Image and Action Recognition
2004.09930 (2020-04)
Relabel the Noise: Joint Extraction of Entities and Relations via Cooperative Multiagents
and 2 more not named here
svm_load_model 1 other paper
1 dated earlier
aaai_25904 (2023 (year only))
Can Label-Specific Features Help Partial-Label Learning?

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 popular technique for sample-efficient optimization of black-box functions. In many applications, the parameters being tuned come with a carefully engineered default configuration, and practitioners only want to deviate from this default when necessary. Standard BO, however, does not aim to minimize deviation from the default and, in practice, often pushes weakly relevant parameters to the boundary of the search space. This makes it difficult to distinguish between important and spurious changes and increases the burden of vetting recommendations when the optimization objective omits relevant operational considerations. We introduce BONSAI, a default-aware BO policy that prunes low-impact deviations from a default configuration while explicitly controlling the loss in acquisition value. BONSAI is compatible with a variety of acquisition functions, including expected improvement and upper confidence bound (GP-UCB). We theoretically bound the regret incurred by BONSAI, showing that, under certain conditions, it enjoys the same no-regret property as vanilla GP-UCB. Moreover, assuming known ARD lengthscales-the same assumption underlying GP-UCB regret bounds-BONSAI provably recovers the relevant-coordinate set at zero acquisition cost, yielding a method that matches the GP-UCB regret rate while recovering the minimal-ℓ 0 solution-a guarantee not provided by prior sparse-BO methods. Across many real-world applications, we empirically find that BONSAI substantially reduces the number of non-default parameters in recommended configurations while maintaining competitive optimization performance, with little effect on wall time-averaging only 1.5× the candidate-generation cost of standard BO, compared to 7-34× on average for prior sparse-BO methods (IR, ER, and SEBO).

For agents

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

get_harvested_code_for_paper("2602.07144")
get_code_for_paper("2602.07144")
have("2602.07144")

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

curl https://syntology.ai/api/ran/2602.07144.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.07144.svg)](https://syntology.ai/paper/2602.07144)

Connect an agent — have() is free.