SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2404.15084 · IJCAI · 2024

Hyperparameter Optimization Can Even Be Harmful in Off-Policy Learning and How to Deal with It

Masahiro Nomura, Yuta Saito

arXiv · PDF · Open in the Atlas

Code that ran

4 of 9 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; 4 failures of our instrument; 5 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/2404.15084.json · badge: /api/ran/2404.15084.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
st-tech/zr-obp canonical (classifier label)
classifier basis: census
4 ran · 5 no recorded run
FunctionStatusWhere it lives
convert_to_action_dist Ran
ran; our contract draft was wrong, not the code
st-tech/zr-obp/obp/utils.py
code served (permissive licence) · get_code("c923dce25aad53dd")
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 3, dtype family float, shape (3, 5, 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

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

# Define parameters
n_actions = 5
n_rounds = 3
len_list = 2

# Create selected_actions array
# Shape: (n_rounds, len_list) with values in range [0, n_actions)
selected_actions = np.random.randint(0, n_actions, size=(n_rounds, len_list))

# Construct ARGS dict
ARGS = {
    "n_actions": n_actions,
    "selected_actions": selected_actions,
}
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 3, dtype family float, shape (3, 5, 2) — shape (3, 5, 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
estimate_confidence_interval_by_bootstrap Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
st-tech/zr-obp/obp/utils.py
code served (permissive licence) · get_code("a1f2bd7d1acda181")
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 numpy as np
from sklearn.utils import check_random_state

# Create sample data for bootstrap confidence interval estimation
np.random.seed(42)
samples = np.random.normal(loc=5.0, scale=2.0, size=100)

# Define arguments for estimate_confidence_interval_by_bootstrap
ARGS = {
    'samples': samples,
    'alpha': 0.05,
    'n_bootstrap_samples': 1000,
    'random_state': 42
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn — 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
sample_action_fast Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
st-tech/zr-obp/obp/utils.py
code served (permissive licence) · get_code("5ae885b38baf7e63")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family int, shape (2,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create a small action distribution matrix
# Shape: (n_rounds=2, n_actions=4)
# Each row should sum to 1.0 for a valid probability distribution
action_dist = np.array([
    [0.1, 0.2, 0.3, 0.4],
    [0.25, 0.25, 0.25, 0.25]
], dtype=np.float64)

ARGS = {
    "action_dist": action_dist,
    "random_state": 42
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family int, shape (2,) — shape (2,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
sample_random_uniform_coefficients Ran
ran; our contract draft was wrong, not the code
st-tech/zr-obp/obp/dataset/synthetic.py
code served (permissive licence) · get_code("4fa7b27a8c130446")
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

ARGS = {
    "effective_dim_action_context": 5,
    "effective_dim_context": 3,
    "random_": np.random.RandomState(42),
}
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
linear_behavior_policy_continuous Not yet run
unverified (harvested, no recorded run)
st-tech/zr-obp/obp/dataset/synthetic_continuous.py
code served (permissive licence) · get_code("0b2a873bf9ce492b")
How it ran

no invocation recorded.

linear_reward_funcion_continuous Not yet run
unverified (harvested, no recorded run)
st-tech/zr-obp/obp/dataset/synthetic_continuous.py
code served (permissive licence) · get_code("8030be7b385db03a")
How it ran

no invocation recorded.

logistic_polynomial_reward_function Not yet run
unverified (harvested, no recorded run)
st-tech/zr-obp/obp/dataset/synthetic.py
code served (permissive licence) · get_code("628f213ec1c901a2")
How it ran

no invocation recorded.

logistic_reward_function Not yet run
unverified (harvested, no recorded run)
st-tech/zr-obp/obp/dataset/synthetic.py
code served (permissive licence) · get_code("99116d6ac9c6fa92")
How it ran

no invocation recorded.

quadratic_reward_funcion_continuous Not yet run
unverified (harvested, no recorded run)
st-tech/zr-obp/obp/dataset/synthetic_continuous.py
code served (permissive licence) · get_code("ad68c45f6b6ffe98")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 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 6 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
convert_to_action_dist 1 other paper
1 dated earlier
2008.07146 (2020-08)
Open Bandit Dataset and Pipeline: Towards Realistic and Reproducible Off-Policy Evaluation
estimate_confidence_interval_by_bootstrap 1 other paper
1 dated earlier
2008.07146 (2020-08)
Open Bandit Dataset and Pipeline: Towards Realistic and Reproducible Off-Policy Evaluation
sample_action_fast 1 other paper
1 dated earlier
2008.07146 (2020-08)
Open Bandit Dataset and Pipeline: Towards Realistic and Reproducible Off-Policy Evaluation

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

There has been a growing interest in off-policy evaluation in the literature such as recommender systems and personalized medicine. We have so far seen significant progress in developing estimators aimed at accurately estimating the effectiveness of counterfactual policies based on biased logged data. However, there are many cases where those estimators are used not only to evaluate the value of decision making policies but also to search for the best hyperparameters from a large candidate space. This work explores the latter hyperparameter optimization (HPO) task for off-policy learning. We empirically show that naively applying an unbiased estimator of the generalization performance as a surrogate objective in HPO can cause an unexpected failure, merely pursuing hyperparameters whose generalization performance is greatly overestimated. We then propose simple and computationally efficient corrections to the typical HPO procedure to deal with the aforementioned issues simultaneously. Empirical investigations demonstrate the effectiveness of our proposed HPO algorithm in situations where the typical procedure fails severely.

For agents

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

get_harvested_code_for_paper("2404.15084")
get_code_for_paper("2404.15084")
have("2404.15084")

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

curl https://syntology.ai/api/ran/2404.15084.json

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

[![Syntology run record](https://syntology.ai/api/ran/2404.15084.svg)](https://syntology.ai/paper/2404.15084)

Connect an agent — have() is free.