The paper proposes a new evaluation test for explanation methods: if an explanation accurately captures how a model uses its features, one should be able to reconstruct the model’s predictions from it. The authors convert explanations into predictors by summing feature effects and assess how well these predictors reproduce the model on unseen data, without any fitting. They apply this test to partial dependence plots, accumulated local effects, SHAP, and LIME across multiple datasets and model families, showing that the best method depends on feature dependence and that some existing quality metrics can favor flawed explanations.
By Jacob Selb{\ae}k, Hugo L. Hammer
arXiv:2606. 06056v1 Announce Type: cross Abstract: Multiple machine learning models can achieve near-equivalent predictive performance on the same task, yet provide divergent feature-based explanations.
By Helge Spieker, J{\o}rn Eirik Betten, Arnaud Gotlieb
arXiv:2605. 27618v2 Announce Type: replace Abstract: Despite the wide use of explainability techniques to attempt to understand the behavior of Artificial Intelligence (AI), the generated explanations may not always be reliable.
By Tom\'as Pereira, Jo\~ao Vitorino, Eva Maia, Isabel Pra\c{c}a
The paper investigates the phenomenon of explanation multiplicity in SHAP, showing that SHAP explanations can vary significantly across repeated runs even when the model, input, and prediction remain unchanged. The authors introduce a dual‑seed evaluation protocol, a hierarchy of metrics, and null models to quantify this variability, finding that multiplicity is widespread across datasets, models, and sampling strategies. They also demonstrate that common metrics like L2 distance may underestimate instability, while rank‑based metrics reveal substantial changes in top‑ranked features, and that improved sampling methods do not fully eliminate run‑to‑run variation.
By Hyunseung Hwang, Seungeun Lee, Lucas Rosenblatt, Steven Euijong Whang, Julia Stoyanovich
The paper introduces MCIR-M, a global feature‑importance method that accounts for feature dependence by conditioning each feature on its strongly dependent neighbours and computing a normalized Mutual Correlation Impact Ratio. MCIR-M yields a population score between 0 and 1, reaching zero when a feature is exactly conditionally redundant, and can be estimated with only a fraction of the data. Experiments on synthetic redundancy setups and the UCI HAR benchmark show that MCIR-M produces more stable rankings under near‑duplicate predictors compared to existing methods such as SHAP, SAGE, HSIC, and MI‑based scores.
By Poushali Sengupta, Sabita Maharjan, Frank Eliassen, Shashi Raj Pandey, Yan Zhang
The paper introduces Adaptive Derivative-Ordered Random Explanation (ADORE), a unified framework that uses first- and second-order derivatives to capture nonlinear feature interactions and feature-sample dynamics. ADORE combines global feature importance with local sample contributions, quantifying both magnitude and direction of feature impact while identifying critical samples. It achieves computational efficiency via randomized SVD and dynamic sparsity detection, outperforming LIME and SHAP across tabular, text, and image data, and is released as an open-source Python package on GitHub.
By Lemen Chao, Ming Lei, Anran Fanga
arXiv:2607. 29614v1 Announce Type: cross Abstract: The rapid adoption of deep learning models in high-risk domains has intensified the need for trustworthy Explainable Artificial Intelligence (XAI).
By Christian Oliva, Luis F. Lago-Fern\'andez
arXiv:2607. 08641v1 Announce Type: new Abstract: Over the last few years, there has been an increased interest in making machine learning models more interpretable.
By Yann Claes, Pierre Geurts, V\^an Anh Huynh-Thu
arXiv:2607. 14315v1 Announce Type: cross Abstract: In this paper, we present a comprehensive framework for assessing the explainability of various XAI methods, such as LIME and SHAP, across multiple datasets and machine learning models, with the ultimate goal of creating a unified multidimensional explainability score.
By Georgios Makridis, Georgios Fatouros, Athanasios Kiourtis, Dimitrios Kotios, Vasileios Koukos, Dimosthenis Kyriazis, Jonh Soldatos
The paper introduces a synthetic ground‑truth framework for evaluating explainable AI (XAI) methods, addressing the lack of reliable evaluation procedures. By using controlled interventions to create datasets where the importance of input components is known, the framework generates ground‑truth explanations that align with the model’s actual decision process. The authors apply this approach to binary images, tabular data, and time series, and find that nine popular XAI methods exhibit significant limitations, underscoring the need for intervention‑based benchmarks.
By Miquel Mir\'o-Nicolau, Francesco Spinnato, Riccardo Guidotti
Background: Understanding individual predictions in complex tabular models remains challenging, as existing additive frameworks like SHAP and LIME often obscure structural logic by absorbing nonlinear...
The paper argues that explainable AI (XAI) tools, intended to increase trust in machine learning models, often fail to do so and instead foster overreliance on incorrect predictions. It attributes this failure to the unfaithfulness of XAI methods, which oversimplify model internals and misalign explanations with the actual model behavior. The authors identify five major shortcomings in current faithfulness metrics and propose four research directions to improve their reliability.
By Marco Zullich, Emily Schiller, Ivan Gentile, David Dembinsky, Adriano Lucieri, Kilian G\"oller, Steffen Seitz