SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2301.11342 · ICML · 2023

A Robust Optimisation Perspective on Counterexample-Guided Repair of Neural Networks

David Boetius, Stefan Leue, Tobias Sutter

arXiv · PDF · Open in the Atlas

Code that ran

3 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; 0 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 2 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/2301.11342.json · badge: /api/ran/2301.11342.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
sen-uni-kn/specrepair canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
get_checkpoint_files_sorted Ran
ran; our contract draft was wrong, not the code
sen-uni-kn/specrepair/scripts/collect_experiment_results.py
code served (permissive licence) · get_code("b8a00efde50c6856")
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.

import re
from pathlib import Path
import tempfile
import os

# Create a temporary directory with mock checkpoint files
temp_dir = tempfile.mkdtemp()
temp_path = Path(temp_dir)

# Create some mock checkpoint files that match a pattern
# Pattern: checkpoint_epoch_<number>_step_<number>.pt
checkpoint_files = [
    "checkpoint_epoch_1_step_100.pt",
    "checkpoint_epoch_2_step_200.pt",
    "checkpoint_epoch_3_step_300.pt",
]

for filename in checkpoint_files:
    (temp_path / filename).touch()

# Also create an ignored file that won't match the pattern
(temp_path / "ignored_file.txt").touch()

# Create a regex pattern that matches checkpoint filenames
# This pattern captures epoch and step numbers as groups
file_name_rx = re.compile(r"checkpoint_epoch_(\d+)_step_(\d+)\.pt")

ARGS = {
    "dir_": temp_path,
    "file_name_rx": file_name_rx,
}
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_runtime_summary Ran
ran; our contract draft was wrong, not the code
sen-uni-kn/specrepair/scripts/collect_experiment_results.py
code served (permissive licence) · get_code("2a4ccb148c642ef5")
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 DataFrame. 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
import pandas
import itertools

# Analyze the function to understand what it needs:
# 1. runtimes_: Dict[str, List[float]] - a dictionary mapping string keys to lists of floats
# 2. prefix_keys: List[str] - a list of prefix strings to filter and aggregate runtimes

# The function:
# - Filters runtimes_ keys using startswith_matching_on_space (key starts with prefix and is either same length or followed by space)
# - Collects matching runtime lists
# - Aggregates statistics (min, mean, median, max, sum) for each prefix
# - Returns a pandas DataFrame

# Construct valid test data
runtimes_dict = {
    'task_a': [1.0, 2.0, 3.0],
    'task_a subcomponent': [0.5, 0.7, 0.9],
    'task_b': [5.0, 6.0],
    'task_b subcomponent': [2.0, 2.5, 3.0],
    'task_c': [10.0, 11.0, 12.0, 13.0],
}

prefix_list = ['task_a', 'task_b', 'task_c']

ARGS = {
    'runtimes_': runtimes_dict,
    'prefix_keys': prefix_list,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
startswith_matching_on_space Ran
ran, violated the contract we drafted
sen-uni-kn/specrepair/scripts/collect_experiment_results.py
code served (permissive licence) · get_code("1d67bdba674223a0")
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 bool, ndim 0, dtype family float, shape () 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

# Analyze the function:
# startswith_matching_on_space(string_: str, prefix_: str)
# This function takes two string parameters and checks if:
# 1. string_ starts with prefix_
# 2. Either string_ and prefix_ have the same length, OR the character after prefix_ in string_ is a space

# Create valid test cases
ARGS = {
    "string_": "hello world",
    "prefix_": "hello"
}
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 bool, ndim 0, dtype family float, shape () — shape () 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 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, 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 0 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_checkpoint_files_sorted 1 other paper
1 dated earlier
2106.01917 (2021-06)
SpecRepair: Counter-Example Guided Safety Repair of Deep Neural Networks
get_runtime_summary 1 other paper
1 dated earlier
2106.01917 (2021-06)
SpecRepair: Counter-Example Guided Safety Repair of Deep Neural Networks
startswith_matching_on_space 1 other paper
1 dated earlier
2106.01917 (2021-06)
SpecRepair: Counter-Example Guided Safety Repair of Deep Neural Networks

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

Counterexample-guided repair aims at creating neural networks with mathematical safety guarantees, facilitating the application of neural networks in safety-critical domains. However, whether counterexample-guided repair is guaranteed to terminate remains an open question. We approach this question by showing that counterexample-guided repair can be viewed as a robust optimisation algorithm. While termination guarantees for neural network repair itself remain beyond our reach, we prove termination for more restrained machine learning models and disprove termination in a general setting. We empirically study the practical implications of our theoretical results, demonstrating the suitability of common verifiers and falsifiers for repair despite a disadvantageous theoretical result. Additionally, we use our theoretical insights to devise a novel algorithm for repairing linear regression models based on quadratic programming, surpassing existing approaches.

For agents

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

get_harvested_code_for_paper("2301.11342")
get_code_for_paper("2301.11342")
have("2301.11342")

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

curl https://syntology.ai/api/ran/2301.11342.json

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

[![Syntology run record](https://syntology.ai/api/ran/2301.11342.svg)](https://syntology.ai/paper/2301.11342)

Connect an agent — have() is free.