Explainable Artificial Intelligence (XAI)

Derniers papiers

• Reasoning Models Will Sometimes Lie About Their Reasoning

  William Walden, Miriam Wanner · 13 avril 2026

  Hint-based faithfulness evaluations have established that Large Reasoning Models (LRMs) may not say what they think: they do not always volunteer information about how key parts of the input (e.g. answer hints) influence their reasoning. Yet, these evaluations also fail to specify what models should…

• Rethinking Prospect Theory for LLMs: Revealing the Instability of Decision-Making under Epistemic Uncertainty

  Rui Wang, Qihan Lin, Jiayu Liu, Qing Zong, Tianshi Zheng, Dadi Guo, Haochen Shi, Weiqi Wang, Yangqiu Song · 13 avril 2026

  Prospect Theory (PT) models human decision-making behaviour under uncertainty, among which linguistic uncertainty is commonly adopted in real-world scenarios. Although recent studies have developed some frameworks to test PT parameters for Large Language Models (LLMs), few have considered the fitnes…

• Uncertainty-Aware Transformers: Conformal Prediction for Language Models

  Abhiram Vellore, Niraj K. Jha · 13 avril 2026

  Transformers have had a profound impact on the field of artificial intelligence, especially on large language models and their variants. However, as was the case with neural networks, their black-box nature limits trust and deployment in high-stakes settings. For models to be genuinely useful and tr…

• Communicate-Predict-Act: Evaluating Social Intelligence of Agents

  David Shoresh, Sarit Kraus, Yonatan Loewenstein · 13 avril 2026

  As large language model (LLM) agents become more prevalent in real world social settings, social intelligence will play an increasingly critical role. But social intelligence is still a poorly defined construct, for humans and artificial agents. We introduce a multiplayer arena of mixed cooperative …

• SPPO: Sequence-Level PPO for Long-Horizon Reasoning Tasks

  Tianyi Wang, Yixia Li, Long Li, Yibiao Chen, Shaohan Huang, Yun Chen, Peng Li, Yang Liu, Guanhua Chen · 13 avril 2026

  Proximal Policy Optimization (PPO) is central to aligning Large Language Models (LLMs) in reasoning tasks with verifiable rewards. However, standard token-level PPO struggles in this setting due to the instability of temporal credit assignment over long Chain-of-Thought (CoT) horizons and the prohib…

• CONDESION-BENCH: Conditional Decision-Making of Large Language Models in Compositional Action Space

  Yeonjun Hwang, Sungyong Park, Minju Kim, Dongha Lee, Jinyoung Yeo · 13 avril 2026

  Large language models have been widely explored as decision-support tools in high-stakes domains due to their contextual understanding and reasoning capabilities. However, existing decision-making benchmarks rely on two simplifying assumptions: actions are selected from a finite set of pre-defined c…

• REVEAL: Reasoning-Enhanced Forensic Evidence Analysis for Explainable AI-Generated Image Detection

  Huangsen Cao, Qin Mei, Zhiheng Li, Yuxi Li, Zhan Meng, Ying Zhang, Chen Li, Zhimeng Zhang, Xin Ding, Yongwei Wang, Jing Lyu, Fei Wu · 10 avril 2026

  The rapid progress of visual generative models has made AI-generated images increasingly difficult to distinguish from authentic ones, posing growing risks to social trust and information integrity. This motivates detectors that are not only accurate but also forensically explainable. While recent m…

• Illocutionary Explanation Planning for Source-Faithful Explanations in Retrieval-Augmented Language Models

  Francesco Sovrano, Alberto Bacchelli · 10 avril 2026

  Natural language explanations produced by large language models (LLMs) are often persuasive, but not necessarily scrutable: users cannot easily verify whether the claims in an explanation are supported by evidence. In XAI, this motivates a focus on faithfulness and traceability, i.e., the extent to …

• Co-design for Trustworthy AI: An Interpretable and Explainable Tool for Type 2 Diabetes Prediction Using Genomic Polygenic Risk Scores

  Ralf Beuthan, Megan Coffee, Heejin Kim, Na Yeon Kim, Pedro Kringen, Elisabeth Hildt, Haekyung Lee, Seunggeun Lee, Emilie Wiinblad Mathez, Sira Maliphol, Vadim Pak, Yuna Park, Stephan Sonnenberg, Jesmin Jahan Tithi, Magnus Westerlund, Roberto V. Zicari · 10 avril 2026

  The polygenic risk scores (PRS) have emerged as an important methodology for quantifying genetic predisposition to complex traits and clinical disease. Significant progress has been made in applying PRS to conditions such as obesity, cancer, and type 2 diabetes (T2DM). Studies have demonstrated that…

• Reasoning Fails Where Step Flow Breaks

  Xiaoyu Xu, Yulan Pan, Xiaosong Yuan, Zhihong Shen, Minghao Su, Yuanhao Su, Xiaofeng Zhang · 10 avril 2026

  Large reasoning models (LRMs) that generate long chains of thought now perform well on multi-step math, science, and coding tasks. However, their behavior is still unstable and hard to interpret, and existing analysis tools struggle with such long, structured reasoning traces. We introduce Step-Sali…

• Non-identifiability of Explanations from Model Behavior in Deep Networks of Image Authenticity Judgments

  Icaro Re Depaolini, Uri Hasson · 9 avril 2026

  Deep neural networks can predict human judgments, but this does not imply that they rely on human-like information or reveal the cues underlying those judgments. Prior work has addressed this issue using attribution heatmaps, but their explanatory value in itself depends on robustness. Here we teste…

• ConceptTracer: Interactive Analysis of Concept Saliency and Selectivity in Neural Representations

  Ricardo Knauer, Andre Beinrucker, Erik Rodner · 9 avril 2026

  Neural networks deliver impressive predictive performance across a variety of tasks, but they are often opaque in their decision-making processes. Despite a growing interest in mechanistic interpretability, tools for systematically exploring the representations learned by neural networks in general,…

• DINO-QPM: Adapting Visual Foundation Models for Globally Interpretable Image Classification

  Robert Zimmermann, Thomas Norrenbrock, Bodo Rosenhahn · 9 avril 2026

  Although visual foundation models like DINOv2 provide state-of-the-art performance as feature extractors, their complex, high-dimensional representations create substantial hurdles for interpretability. This work proposes DINO-QPM, which converts these powerful but entangled features into contrastiv…

• A comparative analysis of machine learning models in SHAP analysis

  Justin Lin, Julia Fukuyama · 9 avril 2026

  In this growing age of data and technology, large black-box models are becoming the norm due to their ability to handle vast amounts of data and learn incredibly complex data patterns. The deficiency of these methods, however, is their inability to explain the prediction process, making them untrust…

• ReDAct: Uncertainty-Aware Deferral for LLM Agents

  Dzianis Piatrashyn, Nikita Kotelevskii, Kirill Grishchenkov, Nikita Glazkov, Ivan Nasonov, Ilya Makarov, Timothy Baldwin, Preslav Nakov, Roman Vashurin, Maxim Panov · 9 avril 2026

  Recently, LLM-based agents have become increasingly popular across many applications, including complex sequential decision-making problems. However, they inherit the tendency of LLMs to hallucinate, leading to incorrect decisions. In sequential settings, even a single mistake can irreversibly degra…

• Improving Robustness In Sparse Autoencoders via Masked Regularization

  Vivek Narayanaswamy, Kowshik Thopalli, Bhavya Kailkhura, Wesam Sakla · 9 avril 2026

  Sparse autoencoders (SAEs) are widely used in mechanistic interpretability to project LLM activations onto sparse latent spaces. However, sparsity alone is an imperfect proxy for interpretability, and current training objectives often result in brittle latent representations. SAEs are known to be pr…

• OmniTabBench: Mapping the Empirical Frontiers of GBDTs, Neural Networks, and Foundation Models for Tabular Data at Scale

  Dihong Jiang, Ruoqi Cao, Zhiyuan Dang, Li Huang, Qingsong Zhang, Zhiyu Wang, Shihao Piao, Shenggao Zhu, Jianlong Chang, Zhouchen Lin, Qi Tian · 9 avril 2026

  While traditional tree-based ensemble methods have long dominated tabular tasks, deep neural networks and emerging foundation models have challenged this primacy, yet no consensus exists on a universally superior paradigm. Existing benchmarks typically contain fewer than 100 datasets, raising concer…

• Emergent Introspection in AI is Content-Agnostic

  Harvey Lederman, Kyle Mahowald · 8 avril 2026

  Introspection is a foundational cognitive ability, but its mechanism is not well understood. Recent work has shown that AI models can introspect. We study the mechanism of this introspection. We first extensively replicate Lindsey (2025)'s thought injection detection paradigm in large open-source mo…

• Pramana: Fine-Tuning Large Language Models for Epistemic Reasoning through Navya-Nyaya

  Sharath Sathish · 8 avril 2026

  Large language models produce fluent text but struggle with systematic reasoning, often hallucinating confident but unfounded claims. When Apple researchers added irrelevant context to mathematical problems, LLM performance degraded by 65% Apple Machine Learning Research, exposing brittle pattern-ma…

• Claw-Eval: Toward Trustworthy Evaluation of Autonomous Agents

  Bowen Ye, Rang Li, Qibin Yang, Yuanxin Liu, Linli Yao, Hanglong Lv, Zhihui Xie, Chenxin An, Lei Li, Lingpeng Kong, Qi Liu, Zhifang Sui, Tong Yang · 8 avril 2026

  Large language models are increasingly deployed as autonomous agents executing multi-step workflows in real-world software environments. However, existing agent benchmarks suffer from three critical limitations: (1) trajectory-opaque grading that checks only final outputs, (2) underspecified safety …

• Weight-Informed Self-Explaining Clustering for Mixed-Type Tabular Data

  Lehao Li, Qiang Huang, Yihao Ang, Bryan Kian Hsiang Low, Anthony K. H. Tung, Xiaokui Xiao · 8 avril 2026

  Clustering mixed-type tabular data is fundamental for exploratory analysis, yet remains challenging due to misaligned numerical-categorical representations, uneven and context-dependent feature relevance, and disconnected and post-hoc explanation from the clustering process. We propose WISE, a Weigh…

• Social Dynamics as Critical Vulnerabilities that Undermine Objective Decision-Making in LLM Collectives

  Changgeon Ko, Jisu Shin, Hoyun Song, Huije Lee, Eui Jun Hwang, Jong C. Park · 8 avril 2026

  Large language model (LLM) agents are increasingly acting as human delegates in multi-agent environments, where a representative agent integrates diverse peer perspectives to make a final decision. Drawing inspiration from social psychology, we investigate how the reliability of this representative …

• Learn to Rank: Visual Attribution by Learning Importance Ranking

  David Schinagl, Christian Fruhwirth-Reisinger, Alexander Prutsch, Samuel Schulter, Horst Possegger · 8 avril 2026

  Interpreting the decisions of complex computer vision models is crucial to establish trust and accountability, especially in safety-critical domains. An established approach to interpretability is generating visual attribution maps that highlight regions of the input most relevant to the model's pre…

• Label Effects: Shared Heuristic Reliance in Trust Assessment by Humans and LLM-as-a-Judge

  Xin Sun, Di Wu, Sijing Qin, Isao Echizen, Abdallah El Ali, Saku Sugawara · 8 avril 2026

  Large language models (LLMs) are increasingly used as automated evaluators (LLM-as-a-Judge). This work challenges its reliability by showing that trust judgments by LLMs are biased by disclosed source labels. Using a counterfactual design, we find that both humans and LLM judges assign higher trust …

• Towards Trustworthy Report Generation: A Deep Research Agent with Progressive Confidence Estimation and Calibration

  Yi Yuan, Xuhong Wang, Shanzhe Lei · 8 avril 2026

  As agent-based systems continue to evolve, deep research agents are capable of automatically generating research-style reports across diverse domains. While these agents promise to streamline information synthesis and knowledge exploration, existing evaluation frameworks-typically based on subjectiv…