SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2010.03039 · 2020

Empirical Frequentist Coverage of Deep Learning Uncertainty Quantification Procedures

arXiv · PDF · Open in the Atlas

Code that ran

6 of 6 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 6 have no contract verdict; 0 failures of our instrument; 0 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/2010.03039.json · badge: /api/ran/2010.03039.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
beamlab-hsph/coverage_quantification community (classifier label)
classifier basis not recorded on the link
6 ran · 0 no recorded run
FunctionStatusWhere it lives
compute_coverage Ran
ran on a synthesized input
beamlab-hsph/coverage_quantification/src/models/find_hyperparameters_non_svi_methods.py
code served (permissive licence) · get_code("5f0acdff8e278043")
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
from sklearn.preprocessing import StandardScaler

# Create a simple normalizer stand-in with inverse_transform method
class SimpleNormalizer:
    def __init__(self, scale=2.0, shift=1.0):
        self.scale = scale
        self.shift = shift
    
    def inverse_transform(self, y):
        return y * self.scale + self.shift

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

# Create sample predictions (2 samples, 4 predictions per sample)
predictions = np.random.randn(2, 4)

# Create a normalizer instance
y_norm = SimpleNormalizer(scale=2.0, shift=1.0)

# Create test labels
y_test = np.array([0.5, 1.0])

ARGS = {
    "predictions": predictions,
    "y_norm": y_norm,
    "y_test": y_test,
}
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 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
compute_quantiles Ran
ran on a synthesized input
beamlab-hsph/coverage_quantification/src/models/find_hyperparameters_non_svi_methods.py
code served (permissive licence) · get_code("4750678b384b6c5b")
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, 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

torch_seed = 0

# Create a small numpy array of predictions
# Shape: (batch_size=2, num_outputs=4)
# The function computes quantiles along axis=1, so we need at least 2 dimensions
predictions = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0]
])

ARGS = {
    "predictions": predictions
}
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, 2) — shape (2, 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_data Ran
ran on a synthesized input
beamlab-hsph/coverage_quantification/src/models/data_utils.py
code served (permissive licence) · get_code("454ea6fd1c88c8fd")
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
import tempfile
import os

# Create temporary directory structure and files for the dataset
temp_dir = tempfile.mkdtemp()
dataset_name = "test_dataset"
data_dir = os.path.join(temp_dir, 'data/DropoutUncertaintyExps/UCI_Datasets', dataset_name, 'data')
os.makedirs(data_dir, exist_ok=True)

# Create mock data files
# Create sample data: 10 samples, 5 features + 1 target = 6 columns
np.random.seed(42)
data = np.random.randn(10, 6)
np.savetxt(os.path.join(data_dir, 'data.txt'), data)

# Feature indices: columns 0, 1, 2, 3 (4 features)
index_features = np.array([0, 1, 2, 3])
np.savetxt(os.path.join(data_dir, 'index_features.txt'), index_features)

# Target index: column 4
index_target = np.array([4])
np.savetxt(os.path.join(data_dir, 'index_target.txt'), index_target)

# Mock the data directory path in the function by patching it
# We'll create a wrapper that modifies the path
original_cwd = os.getcwd()
os.chdir(temp_dir)

ARGS = {
    'dataset': dataset_name
}
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
get_data_file_paths Ran
ran on a synthesized input
beamlab-hsph/coverage_quantification/src/models/data_utils.py
code served (permissive licence) · get_code("98c106d55f5b9346")
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.

# Looking at the function, it only takes a 'dataset' parameter (a string)
# and constructs file paths from it. No file I/O is actually performed,
# just string manipulation to build paths.

ARGS = {
    'dataset': 'boston'
}
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 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
get_index_train_test_path Ran
ran on a synthesized input
beamlab-hsph/coverage_quantification/src/models/data_utils.py
code served (permissive licence) · get_code("eafc9a17384a4762")
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 torch
import numpy as np

# Based on the function body, we need:
# - _DATA_DIRECTORY_PATH: a string representing a directory path
# - split_num: an integer representing the split number (1-20)
# - train: a boolean flag indicating whether to get training or test data

ARGS = {
    '_DATA_DIRECTORY_PATH': '/data/splits/',
    'split_num': 5,
    'train': True
}
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 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
predict_N_times Ran
ran on a synthesized input
beamlab-hsph/coverage_quantification/src/models/find_hyperparameters_non_svi_methods.py
code served (permissive licence) · get_code("7e25fbc2497a51bb")
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 (2, 200, 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
import torch
import torch.nn as nn

torch.manual_seed(0)

# Build a minimal model that has a predict method
class MinimalModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def predict(self, test_set):
        # Simulate prediction: takes test_set and returns predictions
        # test_set should be array-like with shape (batch, features)
        if isinstance(test_set, np.ndarray):
            x = torch.from_numpy(test_set).float()
        else:
            x = test_set
        with torch.no_grad():
            out = self.linear(x)
        return out.numpy()

# Build a minimal normalizer with inverse_transform method
class MinimalNormalizer:
    def __init__(self):
        self.mean = np.array([1.0, 2.0, 3.0])
        self.std = np.array([0.5, 0.5, 0.5])
    
    def inverse_transform(self, y):
        # Denormalize: y_original = y * std + mean
        return y * self.std + self.mean

# Create test_set (batch_size=2, features=4)
test_set = np.random.randn(2, 4).astype(np.float32)

# Instantiate components
model = MinimalModel()
y_norm = MinimalNormalizer()

ARGS = {
    "model": model,
    "test_set": test_set,
    "y_norm": y_norm,
    "num_predictions": 200
}
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 ndarray, ndim 3, dtype family float, shape (2, 200, 3) — shape (2, 200, 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

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 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.

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

Uncertainty quantification for complex deep learning models is increasingly important as these techniques see growing use in high-stakes, real-world settings. Currently, the quality of a model's uncertainty is evaluated using point-prediction metrics such as negative log-likelihood or the Brier score on heldout data. In this study, we provide the first large scale evaluation of the empirical frequentist coverage properties of well known uncertainty quantification techniques on a suite of regression and classification tasks. We find that, in general, some methods do achieve desirable coverage properties on in distribution samples, but that coverage is not maintained on out-of-distribution data. Our results demonstrate the failings of current uncertainty quantification techniques as dataset shift increases and establish coverage as an important metric in developing models for real-world applications.

For agents

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

get_harvested_code_for_paper("2010.03039")
get_code_for_paper("2010.03039")
have("2010.03039")

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

curl https://syntology.ai/api/ran/2010.03039.json

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

[![Syntology run record](https://syntology.ai/api/ran/2010.03039.svg)](https://syntology.ai/paper/2010.03039)

Connect an agent — have() is free.