Deep Learning Reproducibility and Explainable AI (XAI)

A.-M. Leventi-Peetz; T. \"Ostreich

RECORD

Deep Learning Reproducibility and Explainable AI (XAI)

A.-M. Leventi-Peetz, T. Östreich

Machine Learning · Submitted 23 Feb 2022

abstract compiled · 2026-09-23 · no language-model review

The authors' main claim The comparison serves the exploration of the feasibility of creating deterministic, robust DL models and deterministic explainable artificial intelligence (XAI) in practice.

how it is stated · hedged possibility 0.44 read the record →

Comment (0) Read on arXiv ↗

Cite ▾

Claim paper

The compiled record

The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Open the compiled record

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 6,178 input tokens.

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.75 · human participants 0.04 · controlled comparison 0.23 · randomized assignment 0.02 · states quantitative result 0.04 · compares with prior results 0.45 · mentions code or data availability 0.89.

The abstract, sentence by sentence

arXiv listing abstract v3. Results and the main claim are underlined; show every role.

The nondeterminism of Deep Learning (DL) training algorithms and its influence on the explainability of neural network (NN) models are investigated in this work with the help of image classification examples. To discuss the issue, two convolutional neural networks (CNN) have been trained and their results compared. The comparison serves the exploration of the feasibility of creating deterministic, robust DL models and deterministic explainable artificial intelligence (XAI) in practice. Successes and limitation of all here carried out efforts are described in detail. The source code of the attained deterministic models has been listed in this work. Reproducibility is indexed as a development-phase-component of the Model Governance Framework, proposed by the EU within their excellence in AI approach. Furthermore, reproducibility is a requirement for establishing causality for the interpretation of model results and building of trust towards the overwhelming expansion of AI systems applications. Problems that have to be solved on the way to reproducibility and ways to deal with some of them, are examined in this work.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	The nondeterminism of Deep Learning (DL) training algorithms and its influence on the explainability of neural network (NN) models are investigated in this work with the help of image classification examples.	contribution 0.90	0.69	0.07	asserted 0.79
s1	To discuss the issue, two convolutional neural networks (CNN) have been trained and their results compared.	method 0.99	0.94	0.02	asserted 0.91
s2	The comparison serves the exploration of the feasibility of creating deterministic, robust DL models and deterministic explainable artificial intelligence (XAI) in practice.	contribution 0.68	0.32	0.12	hedged possibility 0.44
s3	Successes and limitation of all here carried out efforts are described in detail.	contribution 0.75	0.81	0.06	asserted 0.82
s4	The source code of the attained deterministic models has been listed in this work.	contribution 0.97	0.89	0.05	asserted 0.99
s5	Reproducibility is indexed as a development-phase-component of the Model Governance Framework, proposed by the EU within their excellence in AI approach.	background 1.00	0.10	0.04	asserted 1.00
s6	Furthermore, reproducibility is a requirement for establishing causality for the interpretation of model results and building of trust towards the overwhelming expansion of AI systems applications.	background 0.96	0.06	0.08	asserted 1.00
s7	Problems that have to be solved on the way to reproducibility and ways to deal with some of them, are examined in this work.	contribution 0.95	0.56	0.05	asserted 0.74

What kind of paper the classifier reports this is: empirical study 0.94 · evidence from observation or experiment 0.80.

What was examined, and what was not

sources read

arXiv listing abstract v3

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

The paper

Original abstract

The nondeterminism of Deep Learning (DL) training algorithms and its influence on the explainability of neural network (NN) models are investigated in this work with the help of image classification examples. To discuss the issue, two convolutional neural networks (CNN) have been trained and their results compared. The comparison serves the exploration of the feasibility of creating deterministic, robust DL models and deterministic explainable artificial intelligence (XAI) in practice. Successes and limitation of all here carried out efforts are described in detail. The source code of the attained deterministic models has been listed in this work. Reproducibility is indexed as a development-phase-component of the Model Governance Framework, proposed by the EU within their excellence in AI approach. Furthermore, reproducibility is a requirement for establishing causality for the interpretation of model results and building of trust towards the overwhelming expansion of AI systems applications. Problems that have to be solved on the way to reproducibility and ways to deal with some of them, are examined in this work.

Human review

How Pith checks papers → · Full audit of this paper →

How the literature uses it

3 arXiv papers cite this in 4 sentences

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

Papers citing it · 2 sentences without a confident label

Cites it as support 2 papers

Significance Testing Especially with deep neural networks, even with a fixed set of hyperparameters, performance can be influenced by a number of (stochastic) factors such as the random seed [ref] or even the choice of hardware or framework [ref] .

Experimental Standards for Deep Learning in Natural Language Processing Research · Experiments & Analysis · 0.96
They require traceability to meet regulatory requirements for audit trails and reproducibility [ref] .

X-SYS: A Reference Architecture for Interactive Explanation Systems · XAI System Development · 0.96

What it cites · 8 sentences without a confident label

Cites as support 1 paper

Moreover, DL models and their explanations are vulnerable to errors, for instance when slightest perturbations are introduced to the input image [ref] .

Explaining and Harnessing Adversarial Examples · Conclusions and future work · 0.98

Cites as background 1 paper

A brief but comprehensive history of CNN can be found in many sources, for example in [ref] , whereby the tendency has always been towards making CNN increasingly deeper.

Xception: Deep Learning with Depthwise Separable Convolutions · Network architectures and HW · 0.99

Discussion

Know something about this paper? Leave a comment, or highlight any passage above to comment on it.