Algorithmic explanations are intended to help stakeholders understand opaque algorithmic decisions, but in practice, they often fall short. First, the meaning of algorithmic explanations is often not what one might intuitively expect, so expert knowledge is required to interpret them correctly. Second, recent work has shown that popular explanation algorithms are uninformative about the behavior of complex decision functions. Together, these issues create a gap between what explanations appear to convey and what they actually provide. In this work, we propose Explanation Cards for Explanation Algorithms, which augment standard explanations with complementary information about robustness and validity, as well as clear instructions for interpretation. The complementary information can render otherwise uninformative explanations practically useful, while also helping to detect cases where they are not. Importantly, the interpretation instructions in explanation cards shift responsibility from users to providers: Rather than expecting users to recognize what can and cannot be concluded from an explanation, providers must make this explicit upfront. Using counterfactual explanations and
Post-hoc explanations are widely used to justify, contest, and review automated decisions in high-stakes domains such as lending, employment, and healthcare. Among these methods, SHAP is often treated as providing a reliable account of which features mattered for an individual prediction and is routinely used to support recourse, oversight, and accountability. In practice, however, SHAP explanations can differ substantially across repeated runs, even when the individual, prediction task, and trained model are held fixed. We conceptualize and name this phenomenon explanation multiplicity: the existence of multiple, internally valid but substantively different explanations for the same decision. Explanation multiplicity poses a normative challenge for responsible AI deployment, as it undermines expectations that explanations can reliably identify the reasons for an adverse outcome. We present a comprehensive methodology for characterizing explanation multiplicity in post-hoc feature attribution methods, disentangling sources arising from model training and selection versus stochasticity intrinsic to the explanation pipeline. Furthermore, whether explanation multiplicity is surfaced d
In this paper, we study the problem of AI explanation of misinformation, where the goal is to identify explanation designs that help improve users' misinformation detection abilities and their overall user experiences. Our work is motivated by the limitations of current Explainable AI (XAI) approaches, which predominantly focus on content explanations that elucidate the linguistic features and sentence structures of the misinformation. To address this limitation, we explore various explanations beyond content explanation, such as "social explanation" that considers the broader social context surrounding misinformation, as well as a "combined explanation" where both the content and social explanations are presented in scenarios that are either aligned or misaligned with each other. To evaluate the comparative effectiveness of these AI explanations, we conduct two online crowdsourcing experiments in the COVID-19 (Study 1 on Prolific) and Politics domains (Study 2 on MTurk). Our results show that AI explanations are generally effective in aiding users to detect misinformation, with effectiveness significantly influenced by the alignment between content and social explanations. We also
Large Language Models increasingly rely on self-explanations, such as chain of thought reasoning, to improve performance on multi step question answering. While these explanations enhance accuracy, they are often verbose and costly to generate, raising the question of how much explanation is truly necessary. In this paper, we examine the trade-off between sufficiency, defined as the ability of an explanation to justify the correct answer, and conciseness, defined as the reduction in explanation length. Building on the information bottleneck principle, we conceptualize explanations as compressed representations that retain only the information essential for producing correct answers.To operationalize this view, we introduce an evaluation pipeline that constrains explanation length and assesses sufficiency using multiple language models on the ARC Challenge dataset. To broaden the scope, we conduct experiments in both English, using the original dataset, and Persian, as a resource-limited language through translation. Our experiments show that more concise explanations often remain sufficient, preserving accuracy while substantially reducing explanation length, whereas excessive comp
Explanation mechanisms are increasingly used to support transparency and trust in vision-language models (VLMs), particularly in settings where model decisions require human oversight. However, the robustness of these explanations remains insufficiently understood. In this work, we investigate whether explanation heatmaps in VLMs, particularly CLIP-based models, faithfully reflect model reasoning under adversarial conditions. We show that explanation maps can be systematically manipulated while preserving the model's original prediction, revealing a disconnect between predictive behavior and explanation faithfulness. To study this vulnerability, we introduce X-Shift, a novel grey-box attack that perturbs patch-level visual representations to redirect explanation heatmaps toward semantically irrelevant regions without altering the predicted output. Unlike conventional adversarial attacks that aim to induce misclassification, X-Shift specifically targets the integrity of the explanation process itself. The attack operates without modifying model parameters and generalizes across multiple CLIP architectures and explanation methods. We evaluate the proposed approach on ImageNet-1k, MS-
This scoping review examines the use of student explanation strategies in postsecondary mathematics and statistics education. We analyzed 46 peer-reviewed articles published between 2014 and 2024, categorizing student explanations into three main types: self-explanation, peer explanation and explanation to fictitious others. The review synthesizes the theoretical underpinnings of these strategies, drawing on the retrieval practice hypothesis, generative learning hypothesis, and social presence hypothesis. Our findings indicate that while self-explanation and explaining to fictitious others foster individual cognitive processes enhancing generative thinking, peer explanation have the potential to combine these benefits with collaborative learning. However, explanation to fictitious others have the potential to mitigate some of the negative impacts that may occur in peer explanation, such as more knowledgeable students dominating peer discussions. The efficacy of the methods varies based on implementation, duration, and context. This scoping review contributes to the growing body of literature on generative learning strategies in postsecondary education and provides insights for opti
Explanations on relational data are hard to verify since the explanation structures are more complex (e.g. graphs). To verify interpretable explanations (e.g. explanations of predictions made in images, text, etc.), typically human subjects are used since it does not necessarily require a lot of expertise. However, to verify the quality of a relational explanation requires expertise and is hard to scale-up. GNNExplainer is arguably one of the most popular explanation methods for Graph Neural Networks. In this paper, we develop an approach where we assess the uncertainty in explanations generated by GNNExplainer. Specifically, we ask the explainer to generate explanations for several counterfactual examples. We generate these examples as symmetric approximations of the relational structure in the original data. From these explanations, we learn a factor graph model to quantify uncertainty in an explanation. Our results on several datasets show that our approach can help verify explanations from GNNExplainer by reliably estimating the uncertainty of a relation specified in the explanation.
AI-driven outcomes can be challenging for end-users to understand. Explanations can address two key questions: "Why this outcome?" (factual) and "Why not another?" (counterfactual). While substantial efforts have been made to formalize factual explanations, a precise and comprehensive study of counterfactual explanations is still lacking. This paper proposes a formal definition of counterfactual explanations, proving some properties they satisfy, and examining the relationship with factual explanations. Given that multiple counterfactual explanations generally exist for a specific case, we also introduce a rigorous method to rank these counterfactual explanations, going beyond a simple minimality condition, and to identify the optimal ones. Our experiments with 12 real-world datasets highlight that, in most cases, a single optimal counterfactual explanation emerges. We also demonstrate, via three metrics, that the selected optimal explanation exhibits higher representativeness and can explain a broader range of elements than a random minimal counterfactual. This result highlights the effectiveness of our approach in identifying more robust and comprehensive counterfactual explanati
Despite the fast progress of explanation techniques in modern Deep Neural Networks (DNNs) where the main focus is handling "how to generate the explanations", advanced research questions that examine the quality of the explanation itself (e.g., "whether the explanations are accurate") and improve the explanation quality (e.g., "how to adjust the model to generate more accurate explanations when explanations are inaccurate") are still relatively under-explored. To guide the model toward better explanations, techniques in explanation supervision - which add supervision signals on the model explanation - have started to show promising effects on improving both the generalizability as and intrinsic interpretability of Deep Neural Networks. However, the research on supervising explanations, especially in vision-based applications represented through saliency maps, is in its early stage due to several inherent challenges: 1) inaccuracy of the human explanation annotation boundary, 2) incompleteness of the human explanation annotation region, and 3) inconsistency of the data distribution between human annotation and model explanation maps. To address the challenges, we propose a generic R
Software architecture knowledge transfer is essential for software development, but related documentation is often incomplete or ambiguous, making oral explanations a common means. Our broader aim is to explore how such explanations might be better supported and eventually automated; as a prerequisite, we first investigate how explanations are actually conducted in practice across five areas: explanation topics, explanation plans, supporting artefacts, typical questions, and expectations and challenges. We report on semi-structured interviews with 17 software professionals across diverse organisations and countries. Our findings include that explanations must balance problem- and technical-domain while considering the explainee's role, experience, and goals. Moreover, explainees frequently seek not only structure and behaviour, but also decision rationale. We propose the Explanation Window, a framework for focusing information by adjusting functionality scope and detail, and emphasise the importance of including system context. These findings provide an empirical basis for improving architecture explanations and guiding future work on tool support and automation.
This paper develops a formal account of causal explanation, grounded in a theory of conversational pragmatics, and inspired by the interventionist idea that explanation is about asking and answering what-if-things-had-been-different questions. We illustrate the fruitfulness of the account, relative to previous accounts, by showing that widely recognised explanatory virtues emerge naturally, as do subtle empirical patterns concerning the impact of norms on causal judgments. This shows the value of a communication-first approach to explanation: getting clear on explanation's communicative dimension is an important prerequisite for philosophical work on explanation. The result is a simple but powerful framework for incorporating insights from the cognitive sciences into philosophical work on explanation, which will be useful for philosophers or cognitive scientists interested in explanation.
Explanations for machine learning models can be hard to interpret or be wrong. Combining an explanation method with an uncertainty estimation method produces explanation uncertainty. Evaluating explanation uncertainty is difficult. In this paper we propose sanity checks for uncertainty explanation methods, where a weight and data randomization tests are defined for explanations with uncertainty, allowing for quick tests to combinations of uncertainty and explanation methods. We experimentally show the validity and effectiveness of these tests on the CIFAR10 and California Housing datasets, noting that Ensembles seem to consistently pass both tests with Guided Backpropagation, Integrated Gradients, and LIME explanations.
Explanations of model behavior are commonly evaluated via proxy properties weakly tied to the purposes explanations serve in practice. We contribute a decision theoretic framework that treats explanations as information signals valued by the expected improvement they enable on a specified decision task. This approach yields three distinct estimands: 1) a theoretical benchmark that upperbounds achievable performance by any agent with the explanation, 2) a human-complementary value that quantifies the theoretically attainable value that is not already captured by a baseline human decision policy, and 3) a behavioral value representing the causal effect of providing the explanation to human decision-makers. We instantiate these definitions in a practical validation workflow, and apply them to assess explanation potential and interpret behavioral effects in human-AI decision support and mechanistic interpretability.
Recent concept-based interpretable models have succeeded in providing meaningful explanations by pre-defined concept sets. However, the dependency on the pre-defined concepts restricts the application because of the limited number of concepts for explanations. This paper proposes a novel interpretable deep neural network called explanation bottleneck models (XBMs). XBMs generate a text explanation from the input without pre-defined concepts and then predict a final task prediction based on the generated explanation by leveraging pre-trained vision-language encoder-decoder models. To achieve both the target task performance and the explanation quality, we train XBMs through the target task loss with the regularization penalizing the explanation decoder via the distillation from the frozen pre-trained decoder. Our experiments, including a comparison to state-of-the-art concept bottleneck models, confirm that XBMs provide accurate and fluent natural language explanations without pre-defined concept sets. Code is available at https://github.com/yshinya6/xbm/.
Code reviews are central for software quality assurance. Ideally, reviewers should explain their feedback to enable authors of code changes to understand the feedback and act accordingly. Different developers might need different explanations in different contexts. Therefore, assisting this process first requires understanding the types of explanations reviewers usually provide. The goal of this paper is to study the types of explanations used in code reviews and explore the potential of Large Language Models (LLMs), specifically ChatGPT, in generating these specific types. We extracted 793 code review comments from Gerrit and manually labeled them based on whether they contained a suggestion, an explanation, or both. Our analysis shows that 42% of comments only include suggestions without explanations. We categorized the explanations into seven distinct types including rule or principle, similar examples, and future implications. When measuring their prevalence, we observed that some explanations are used differently by novice and experienced reviewers. Our manual evaluation shows that, when the explanation type is specified, ChatGPT can correctly generate the explanation in 88 ou
Why do we give the explanations we do? Recent work has suggested that we should think of explanation as a kind of cooperative social interaction, between a why-question-asker and an explainer. Here, we apply this perspective to consider the role that emotion plays in this social interaction. We develop a computational framework for modeling explainers who consider the emotional impact an explanation might have on a listener. We test our framework by using it to model human intuitions about how a doctor might explain to a patient why they have a disease, taking into account the patient's propensity for regret. Our model predicts human intuitions well, better than emotion-agnostic ablations, suggesting that people do indeed reason about emotion when giving explanations.
Generating explanation to explain its behavior is an essential capability for a robotic teammate. Explanations help human partners better understand the situation and maintain trust of their teammates. Prior work on robot generating explanations focuses on providing the reasoning behind its decision making. These approaches, however, fail to heed the cognitive requirement of understanding an explanation. In other words, while they provide the right explanations from the explainer's perspective, the explainee part of the equation is ignored. In this work, we address an important aspect along this direction that contributes to a better understanding of a given explanation, which we refer to as the progressiveness of explanations. A progressive explanation improves understanding by limiting the cognitive effort required at each step of making the explanation. As a result, such explanations are expected to be smoother and hence easier to understand. A general formulation of progressive explanation is presented. Algorithms are provided based on several alternative quantifications of cognitive effort as an explanation is being made, which are evaluated in a standard planning competition
Fairness-aware recommendation eliminates discrimination issues to build trustworthy recommendation systems.Explaining the causes of unfair recommendations is critical, as it promotes fairness diagnostics, and thus secures users' trust in recommendation models. Existing fairness explanation methods suffer high computation burdens due to the large-scale search space and the greedy nature of the explanation search process. Besides, they perform score-based optimizations with continuous values, which are not applicable to discrete attributes such as gender and race. In this work, we adopt the novel paradigm of counterfactual explanation from causal inference to explore how minimal alterations in explanations change model fairness, to abandon the greedy search for explanations. We use real-world attributes from Heterogeneous Information Networks (HINs) to empower counterfactual reasoning on discrete attributes. We propose a novel Counterfactual Explanation for Fairness (CFairER) that generates attribute-level counterfactual explanations from HINs for recommendation fairness. Our CFairER conducts off-policy reinforcement learning to seek high-quality counterfactual explanations, with an
Text-based explanation is a particularly promising approach in explainable AI, but the evaluation of text explanations is method-dependent. We argue that placing the explanations on an information-theoretic framework could unify the evaluations of two popular text explanation methods: rationale and natural language explanations (NLE). This framework considers the post-hoc text pipeline as a series of communication channels, which we refer to as ``explanation channels''. We quantify the information flow through these channels, thereby facilitating the assessment of explanation characteristics. We set up tools for quantifying two information scores: relevance and informativeness. We illustrate what our proposed information scores measure by comparing them against some traditional evaluation metrics. Our information-theoretic scores reveal some unique observations about the underlying mechanisms of two representative text explanations. For example, the NLEs trade-off slightly between transmitting the input-related information and the target-related information, whereas the rationales do not exhibit such a trade-off mechanism. Our work contributes to the ongoing efforts in establishing
As larger deep learning models are hard to interpret, there has been a recent focus on generating explanations of these black-box models. In contrast, we may have apriori explanations of how models should behave. In this paper, we formalize this notion as learning from explanation constraints and provide a learning theoretic framework to analyze how such explanations can improve the learning of our models. One may naturally ask, "When would these explanations be helpful?" Our first key contribution addresses this question via a class of models that satisfies these explanation constraints in expectation over new data. We provide a characterization of the benefits of these models (in terms of the reduction of their Rademacher complexities) for a canonical class of explanations given by gradient information in the settings of both linear models and two layer neural networks. In addition, we provide an algorithmic solution for our framework, via a variational approximation that achieves better performance and satisfies these constraints more frequently, when compared to simpler augmented Lagrangian methods to incorporate these explanations. We demonstrate the benefits of our approach o