Explainable Artificial Intelligence (XAI)

Derniers papiers

• SCOPE: Selective Conformal Optimized Pairwise LLM Judging

  Sher Badshah, Ali Emami, Hassan Sajjad · 16 février 2026

  Large language models (LLMs) are increasingly used as judges to replace costly human preference labels in pairwise evaluation. Despite their practicality, LLM judges remain prone to miscalibration and systematic biases. This paper proposes SCOPE (Selective Conformal Optimized Pairwise Evaluation), a…

• A Theoretical Framework for Adaptive Utility-Weighted Benchmarking

  Philip Waggoner · 16 février 2026

  Benchmarking has long served as a foundational practice in machine learning and, increasingly, in modern AI systems such as large language models, where shared tasks, metrics, and leaderboards offer a common basis for measuring progress and comparing approaches. As AI systems are deployed in more va…

• X-SYS: A Reference Architecture for Interactive Explanation Systems

  Tobias Labarta, Nhi Hoang, Maximilian Dreyer, Jim Berend, Oleg Hein, Jackie Ma, Wojciech Samek, Sebastian Lapuschkin · 16 février 2026

  The explainable AI (XAI) research community has proposed numerous technical methods, yet deploying explainability as systems remains challenging: Interactive explanation systems require both suitable algorithms and system capabilities that maintain explanation usability across repeated queries, evol…

• Uncertainty in Federated Granger Causality: From Origins to Systemic Consequences

  Ayush Mohanty, Nazal Mohamed, Nagi Gebraeel · 16 février 2026

  Granger Causality (GC) provides a rigorous framework for learning causal structures from time-series data. Recent federated variants of GC have targeted distributed infrastructure applications (e.g., smart grids) with distributed clients that generate high-dimensional data bound by data-sovereignty …

• Aggregate Models, Not Explanations: Improving Feature Importance Estimation

  Joseph Paillard, Angel Reyero Lobo, Denis A. Engemann, Bertrand Thirion · 13 février 2026

  Feature-importance methods show promise in transforming machine learning models from predictive engines into tools for scientific discovery. However, due to data sampling and algorithmic stochasticity, expressive models can be unstable, leading to inaccurate variable importance estimates and undermi…

• How Sampling Shapes LLM Alignment: From One-Shot Optima to Iterative Dynamics

  Yurong Chen, Yu He, Michael I. Jordan, Fan Yao · 13 février 2026

  Standard methods for aligning large language models with human preferences learn from pairwise comparisons among sampled candidate responses and regularize toward a reference policy. Despite their effectiveness, the effects of sampling and reference choices are poorly understood theoretically. We in…

• Improving Neural Retrieval with Attribution-Guided Query Rewriting

  Moncef Garouani, Josiane Mothe · 13 février 2026

  Neural retrievers are effective but brittle: underspecified or ambiguous queries can misdirect ranking even when relevant documents exist. Existing approaches address this brittleness only partially: LLMs rewrite queries without retriever feedback, and explainability methods identify misleading toke…

• STAR : Bridging Statistical and Agentic Reasoning for Large Model Performance Prediction

  Xiaoxiao Wang, Chunxiao Li, Junying Wang, Yijin Guo, Zijian Chen, Chunyi Li, Xiaohong Liu, Zicheng Zhang, Guangtao Zhai · 13 février 2026

  As comprehensive large model evaluation becomes prohibitively expensive, predicting model performance from limited observations has become essential. However, existing statistical methods struggle with pattern shifts, data sparsity, and lack of explanation, while pure LLM methods remain unreliable. …

• Locally Interpretable Individualized Treatment Rules for Black-Box Decision Models

  Yasin Khadem Charvadeh, Katherine S. Panageas, Yuan Chen · 13 février 2026

  Individualized treatment rules (ITRs) aim to optimize healthcare by tailoring treatment decisions to patient-specific characteristics. Existing methods typically rely on either interpretable but inflexible models or highly flexible black-box approaches that sacrifice interpretability; moreover, most…

• Credal Concept Bottleneck Models: Structural Separation of Epistemic and Aleatoric Uncertainty

  Tanmoy Mukherjee, Marius Kloft, Pierre Marquis, Zied Bouraoui · 13 février 2026

  Decomposing predictive uncertainty into epistemic (model ignorance) and aleatoric (data ambiguity) components is central to reliable decision making, yet most methods estimate both from the same predictive distribution. Recent empirical and theoretical results show these estimates are typically stro…

• Control Reinforcement Learning: Interpretable Token-Level Steering of LLMs via Sparse Autoencoder Features

  Seonglae Cho, Zekun Wu, Adriano Koshiyama · 13 février 2026

  Sparse autoencoders (SAEs) decompose language model activations into interpretable features, but existing methods reveal only which features activate, not which change model outputs when amplified. We introduce Control Reinforcement Learning (CRL), which trains a policy to select SAE features for st…

• Leveraging Generative AI for Human Understanding: Meta-Requirements and Design Principles for Explanatory AI as a new Paradigm

  Christian Meske, Justin Brenne, Erdi Uenal, Sabahat Oelcer, Ayseguel Doganguen · 13 février 2026

  Artificial intelligence (AI) systems increasingly support decision-making across critical domains, yet current explainable AI (XAI) approaches prioritize algorithmic transparency over human comprehension. While XAI methods reveal computational processes for model validation and audit, end users requ…

• Hierarchical Concept Embedding & Pursuit for Interpretable Image Classification

  Nghia Nguyen, Tianjiao Ding, Ren\'e Vidal · 13 février 2026

  Interpretable-by-design models are gaining traction in computer vision because they provide faithful explanations for their predictions. In image classification, these models typically recover human-interpretable concepts from an image and use them for classification. Sparse concept recovery methods…

• TreeGrad-Ranker: Feature Ranking via $O(L)$-Time Gradients for Decision Trees

  Weida Li, Yaoliang Yu, Bryan Kian Hsiang Low · 13 février 2026

  We revisit the use of probabilistic values, which include the well-known Shapley and Banzhaf values, to rank features for explaining the local predicted values of decision trees. The quality of feature rankings is typically assessed with the insertion and deletion metrics. Empirically, we observe th…

• Benchmark Health Index: A Systematic Framework for Benchmarking the Benchmarks of LLMs

  Longyuan Zhu, Hairan Hua, Linlin Miao, Bing Zhao · 13 février 2026

  Large Language Models (LLMs) are advancing rapidly, yet the benchmarks used to measure this progress are becoming increasingly unreliable. Score inflation and selective reporting have eroded the authority of standard benchmarks, leaving the community uncertain about which evaluation results remain t…

• Calibrated Bayesian Deep Learning for Explainable Decision Support Systems Based on Medical Imaging

  Hua Xu, Juli\'an D. Arias-Londo\~no, Juan I. Godino-Llorente · 13 février 2026

  In critical decision support systems based on medical imaging, the reliability of AI-assisted decision-making is as relevant as predictive accuracy. Although deep learning models have demonstrated significant accuracy, they frequently suffer from miscalibration, manifested as overconfidence in erron…

• AgentNoiseBench: Benchmarking Robustness of Tool-Using LLM Agents Under Noisy Condition

  Ruipeng Wang, Yuxin Chen, Yukai Wang, Chang Wu, Junfeng Fang, Xiaodong Cai, Qi Gu, Hui Su, An Zhang, Xiang Wang, Xunliang Cai, Tat-Seng Chua · 13 février 2026

  Recent advances in large language models have enabled LLM-based agents to achieve strong performance on a variety of benchmarks. However, their performance in real-world deployments often that observed on benchmark settings, especially in complex and imperfect environments. This discrepancy largely …

• Right for the Wrong Reasons: Epistemic Regret Minimization for Causal Rung Collapse in LLMs

  Edward Y. Chang · 13 février 2026

  Machine learning systems that are "right for the wrong reasons" achieve high performance through shortcuts that collapse under distributional shift. We show this pathology has a precise causal origin: autoregressive training provides no gradient signal to distinguish association P(Y|X) from interven…

• Explaining AI Without Code: A User Study on Explainable AI

  Natalia Abarca, Andr\'es Carvallo, Claudia L\'opez Moncada, Felipe Bravo-Marquez · 13 février 2026

  The increasing use of Machine Learning (ML) in sensitive domains such as healthcare, finance, and public policy has raised concerns about the transparency of automated decisions. Explainable AI (XAI) addresses this by clarifying how models generate predictions, yet most methods demand technical expe…

• CausalAgent: A Conversational Multi-Agent System for End-to-End Causal Inference

  Jiawei Zhu, Wei Chen, Ruichu Cai · 13 février 2026

  Causal inference holds immense value in fields such as healthcare, economics, and social sciences. However, traditional causal analysis workflows impose significant technical barriers, requiring researchers to possess dual backgrounds in statistics and computer science, while manually selecting algo…

• GRASP: group-Shapley feature selection for patients

  Yuheng Luo, Shuyan Li, Zhong Cao · 12 février 2026

  Feature selection remains a major challenge in medical prediction, where existing approaches such as LASSO often lack robustness and interpretability. We introduce GRASP, a novel framework that couples Shapley value driven attribution with group $L_{21}$ regularization to extract compact and non-red…

• Analyzing Fairness of Neural Network Prediction via Counterfactual Dataset Generation

  Brian Hyeongseok Kim, Jacqueline L. Mitchell, Chao Wang · 12 février 2026

  Interpreting the inference-time behavior of deep neural networks remains a challenging problem. Existing approaches to counterfactual explanation typically ask: What is the closest alternative input that would alter the model's prediction in a desired way? In contrast, we explore counterfactual data…

• Statistical Inference and Learning for Shapley Additive Explanations (SHAP)

  Justin Whitehouse, Ayush Sawarni, Vasilis Syrgkanis · 12 février 2026

  The SHAP (short for Shapley additive explanation) framework has become an essential tool for attributing importance to variables in predictive tasks. In model-agnostic settings, SHAP uses the concept of Shapley values from cooperative game theory to fairly allocate credit to the features in a vector…

• Control Reinforcement Learning: Token-Level Mechanistic Analysis via Learned SAE Feature Steering

  Seonglae Cho, Zekun Wu, Adriano Koshiyama · 12 février 2026

  Sparse autoencoders (SAEs) decompose language model activations into interpretable features, but existing methods reveal only which features activate, not which change model outputs when amplified. We introduce Control Reinforcement Learning (CRL), which trains a policy to select SAE features for st…

• The Alignment Bottleneck in Decomposition-Based Claim Verification

  Mahmud Elahi Akhter, Federico Ruggeri, Iman Munire Bilal, Rob Procter, Maria Liakata · 12 février 2026

  Structured claim decomposition is often proposed as a solution for verifying complex, multi-faceted claims, yet empirical results have been inconsistent. We argue that these inconsistencies stem from two overlooked bottlenecks: evidence alignment and sub-claim error profiles. To better understand th…