SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.08129 · ICML · 2020

Bayesian Experimental Design for Implicit Models by Mutual Information Neural Estimation

Steven Kleinegesse, Michael Gutmann

arXiv · PDF · Open in the Atlas

Code that ran

2 of 3 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure of our instrument; 1 no recorded run), measured 2026-11-21; 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/2002.08129.json · badge: /api/ran/2002.08129.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
stevenkleinegesse/minebed canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
FunctionStatusWhere it lives
fun_dfun Ran
ran; our contract draft was wrong, not the code
stevenkleinegesse/minebed/src/minebed/methods.py
code served (permissive licence) · get_code("ed9d6f8206e51233")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned tuple. The image it ran in was not recorded. 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 unittest.mock import Mock, MagicMock

# Create a mock GPyOpt model that supports predict_withGradients
mock_model = Mock()

# predict_withGradients returns a tuple of (predictions, uncertainties, gradients)
# We need to return values that match the indexing in the function:
# pred = obj.model.predict_withGradients(d)[0][0][0]
# d_pred = obj.model.predict_withGradients(d)[2][0]

predictions = np.array([[5.0]])  # Shape (1, 1) for [0][0][0] indexing
uncertainties = np.array([[0.5]])  # Not used but part of standard return
gradients = np.array([[0.1, 0.2, 0.3, 0.4]])  # Shape (1, 4) for [2][0] indexing

mock_model.predict_withGradients.return_value = (
    predictions,
    uncertainties,
    gradients
)

# Create a mock GPyOpt object with the model
obj = Mock()
obj.model = mock_model

# Create a mock GPyOpt space with indicator_constraints method
space = Mock()

# indicator_constraints should return a mask (scalar or array)
# Based on usage: float(pred * mask), it should be a scalar or broadcastable
space.indicator_constraints.return_value = 1.0

# Create a design vector d with shape (1, 4)
d = np.array([[0.5, 0.3, 0.7, 0.2]])

ARGS = {
    "obj": obj,
    "space": space,
    "d": d
}
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 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
indicator_boundaries Ran
ran, honoured the contract we drafted
stevenkleinegesse/minebed/src/minebed/methods.py
code served (permissive licence) · get_code("6f1dad4692c9062d")
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 ndarray, ndim 2, dtype family float, shape (1, 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 numpy as np

# The function takes two parameters:
# 1. bounds: np.ndarray - should be a 2D array where bounds.T has shape (2, d)
#    meaning bounds has shape (d, 2) with lower and upper bounds for each dimension
# 2. d: np.ndarray - a 1D design array that will be compared against bounds

# Create a small design space with 4 dimensions
# bounds should have shape (4, 2) - 4 dimensions, each with [lower, upper]
bounds = np.array([
    [0.0, 1.0],      # dimension 0: bounds from 0 to 1
    [0.0, 2.0],      # dimension 1: bounds from 0 to 2
    [-1.0, 1.0],     # dimension 2: bounds from -1 to 1
    [0.5, 3.5]       # dimension 3: bounds from 0.5 to 3.5
])

# Create a design point that is within the bounds
# This should return ind = 1.0
d = np.array([0.5, 1.0, 0.0, 2.0])

ARGS = {
    "bounds": bounds,
    "d": d
}
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 ndarray, ndim 2, dtype family float, shape (1, 1) — shape (1, 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
simulator Not yet run
unverified (harvested, no recorded run)
stevenkleinegesse/minebed/examples/bed_bo_example.py
code served (permissive licence) · get_code("d99b97f19d4e9b32")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 2 later) and 1 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
fun_dfun 2 other papers
1 dated earlier, 1 later
1810.09912 (2018-10)
Efficient Bayesian Experimental Design for Implicit Models
2003.09379 (2020-03)
Sequential Bayesian Experimental Design for Implicit Models via Mutual Information
indicator_boundaries 2 other papers
1 dated earlier, 1 later
1810.09912 (2018-10)
Efficient Bayesian Experimental Design for Implicit Models
2003.09379 (2020-03)
Sequential Bayesian Experimental Design for Implicit Models via Mutual Information

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 2026-11-21.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Implicit stochastic models, where the datageneration distribution is intractable but sampling is possible, are ubiquitous in the natural sciences. The models typically have free parameters that need to be inferred from data collected in scientific experiments. A fundamental question is how to design the experiments so that the collected data are most useful. The field of Bayesian experimental design advocates that, ideally, we should choose designs that maximise the mutual information (MI) between the data and the parameters. For implicit models, however, this approach is severely hampered by the high computational cost of computing posteriors and maximising MI, in particular when we have more than a handful of design variables to optimise. In this paper, we propose a new approach to Bayesian experimental design for implicit models that leverages recent advances in neural MI estimation to deal with these issues. We show that training a neural network to maximise a lower bound on MI allows us to jointly determine the optimal design and the posterior. Simulation studies illustrate that this gracefully extends Bayesian experimental design for implicit models to higher design dimensions.

For agents

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

get_harvested_code_for_paper("2002.08129")
get_code_for_paper("2002.08129")
have("2002.08129")

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

curl https://syntology.ai/api/ran/2002.08129.json

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

[![Syntology run record](https://syntology.ai/api/ran/2002.08129.svg)](https://syntology.ai/paper/2002.08129)

Connect an agent — have() is free.