<?xml version="1.0" encoding="utf-8"?>
<export-example>
  <doc>
    <id>4952</id>
    <completedYear/>
    <publishedYear>2023</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume>abs/2308.06100</volume>
    <type>article</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>1</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Diffusion-based Visual Counterfactual Explanations - Towards Systematic Quantitative Evaluation</title>
    <abstract language="eng">Latest methods for visual counterfactual explanations (VCE) harness the power of deep generative models to synthesize new examples of high-dimensional images of impressive quality. However, it is currently difficult to compare the performance of these VCE methods as the evaluation procedures largely vary and often boil down to visual inspection of individual examples and small scale user studies. In this work, we propose a framework for systematic, quantitative evaluation of the VCE methods and a minimal set of metrics to be used. We use this framework to explore the effects of certain crucial design choices in the latest diffusion-based generative models for VCEs of natural image classification (ImageNet). We conduct a battery of ablation-like experiments, generating thousands of VCEs for a suite of classifiers of various complexity, accuracy and robustness. Our findings suggest multiple directions for future advancements and improvements of VCE methods. By sharing our methodology and our approach to tackle the computational challenges of such a study on a limited hardware setup (including the complete code base), we offer a valuable guidance for researchers in the field fostering consistency and transparency in the assessment of counterfactual explanations.</abstract>
    <parentTitle language="eng">CoRR</parentTitle>
    <identifier type="doi">10.48550/arXiv.2308.06100</identifier>
    <enrichment key="opus.import.data">@articleDBLP:journals/corr/abs-2308-06100, author = Philipp Väth and Alexander M. Frühwald and Benjamin Paaßen and Magda Gregorova, title = Diffusion-based Visual Counterfactual Explanations - Towards Systematic Quantitative Evaluation, journal = CoRR, volume = abs/2308.06100, year = 2023, url = https://doi.org/10.48550/arXiv.2308.06100, doi = 10.48550/arXiv.2308.06100, eprinttype = arXiv, eprint = 2308.06100, timestamp = Wed, 23 Aug 2023 01:00:00 +0200, biburl = https://dblp.org/rec/journals/corr/abs-2308-06100.bib, bibsource = dblp computer science bibliography, https://dblp.org</enrichment>
    <enrichment key="opus.import.dataHash">md5:78e7efeccb5fd5a81999128a4cf76812</enrichment>
    <enrichment key="opus.import.date">2023-12-05T10:39:47+00:00</enrichment>
    <enrichment key="opus.import.file">/tmp/phpbqftxu</enrichment>
    <enrichment key="opus.import.format">bibtex</enrichment>
    <enrichment key="opus.import.id">656efdf3f01634.07906115</enrichment>
    <author>Philipp Väth</author>
    <author>Alexander M. Frühwald</author>
    <author>Benjamin Paaßen</author>
    <author>Magda Gregorová</author>
    <collection role="institutes" number="fiw">Fakultät Informatik und Wirtschaftsinformatik</collection>
    <collection role="institutes" number="cairo">Center for Artificial Intelligence (CAIRO)</collection>
  </doc>
  <doc>
    <id>5918</id>
    <completedYear/>
    <publishedYear>2024</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume>abs/2406.17399</volume>
    <type>article</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>1</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">GradCheck: Analyzing classifier guidance gradients for conditional diffusion sampling</title>
    <parentTitle language="eng">CoRR</parentTitle>
    <identifier type="doi">10.48550/ARXIV.2406.17399</identifier>
    <enrichment key="opus.import.data">@articleDBLP:journals/corr/abs-2406-17399, author = Philipp Väth and Alexander M. Frühwald and Benjamin Paassen and Magda Gregorova, title = GradCheck: Analyzing classifier guidance gradients for conditional diffusion sampling, journal = CoRR, volume = abs/2406.17399, year = 2024, url = https://doi.org/10.48550/arXiv.2406.17399, doi = 10.48550/ARXIV.2406.17399, eprinttype = arXiv, eprint = 2406.17399, timestamp = Tue, 23 Jul 2024 01:00:00 +0200, biburl = https://dblp.org/rec/journals/corr/abs-2406-17399.bib, bibsource = dblp computer science bibliography, https://dblp.org</enrichment>
    <enrichment key="opus.import.dataHash">md5:557b2daa1b9eca66288cfa4805cad8aa</enrichment>
    <enrichment key="opus.import.date">2025-07-02T12:42:53+00:00</enrichment>
    <enrichment key="opus.import.file">/tmp/phpQJldBV</enrichment>
    <enrichment key="opus.import.format">bibtex</enrichment>
    <enrichment key="opus.import.id">6865294dca6e23.97476895</enrichment>
    <author>Philipp Väth</author>
    <author>Alexander M. Frühwald</author>
    <author>Benjamin Paassen</author>
    <author>Magda Gregorová</author>
    <collection role="institutes" number="fiw">Fakultät Informatik und Wirtschaftsinformatik</collection>
    <collection role="institutes" number="cairo">Center for Artificial Intelligence (CAIRO)</collection>
  </doc>
  <doc>
    <id>5919</id>
    <completedYear/>
    <publishedYear>2024</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume>abs/2410.20890</volume>
    <type>article</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>1</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Generative Example-Based Explanations: Bridging the Gap between Generative Modeling and Explainability</title>
    <parentTitle language="eng">CoRR</parentTitle>
    <identifier type="doi">10.48550/ARXIV.2410.20890</identifier>
    <enrichment key="opus.import.data">@articleDBLP:journals/corr/abs-2410-20890, author = Philipp Väth and Alexander M. Frühwald and Benjamin Paassen and Magda Gregorova, title = Generative Example-Based Explanations: Bridging the Gap between Generative Modeling and Explainability, journal = CoRR, volume = abs/2410.20890, year = 2024, url = https://doi.org/10.48550/arXiv.2410.20890, doi = 10.48550/ARXIV.2410.20890, eprinttype = arXiv, eprint = 2410.20890, timestamp = Fri, 29 Nov 2024 00:00:00 +0100, biburl = https://dblp.org/rec/journals/corr/abs-2410-20890.bib, bibsource = dblp computer science bibliography, https://dblp.org</enrichment>
    <enrichment key="opus.import.dataHash">md5:035adfc1d8ca1d545b3eae120afbbaea</enrichment>
    <enrichment key="opus.import.date">2025-07-02T12:42:53+00:00</enrichment>
    <enrichment key="opus.import.file">/tmp/phpQJldBV</enrichment>
    <enrichment key="opus.import.format">bibtex</enrichment>
    <enrichment key="opus.import.id">6865294dca6e23.97476895</enrichment>
    <author>Philipp Väth</author>
    <author>Alexander M. Frühwald</author>
    <author>Benjamin Paassen</author>
    <author>Magda Gregorová</author>
    <collection role="institutes" number="fiw">Fakultät Informatik und Wirtschaftsinformatik</collection>
    <collection role="institutes" number="cairo">Center for Artificial Intelligence (CAIRO)</collection>
  </doc>
</export-example>
