Explainable Artificial Intelligence (XAI)

Derniers papiers

• Workflow-GYM: Towards Long-Horizon Evaluation of Computer-use Agentic tasks in Real-World Professional Fields

  Liya Zhu, Jingzhe Ding, Jian Zhang, Jianbo Xue, Shihao Liang, Ge Zhang, Xiang Gao, Qingshui Gu, Mailun Gao, Huimin Che, Yan Zhao, Peiheng Zhou, Haojun Wang, Chaobo Xian, Lili Le, Chi Wu, Yiwei Liu, Shengda Long, Jiale Yang, Fangzhi Xu, Sijin Wu, Haodong Duan, Yi Zhu, Chao He, Zhaojian Li, Minchao Wang, Huan Zhou, Jiani Hou, Chuqian Yu, Weiran Shi, Hongwan Gao, Jiamin Chen, Guanhong Chen, Tingqin Luo, Kaiyuan Zhang, Zhixin Yao, Qing Hua, Yuhao Jiang, Jin Chen, Pu Chen, Zhenyu Hu, Xingyu Li, Zhengxuan Jiang, Meng Cao, Tianfeng Long, Haozhe Wang, Mingzhang Wang, Yichen Zhang, Yiming Dai, Chenchen Zhang, Jiaying Wang, Zhiyong Wu, Shen Yan, Yujia Qin, Wenhao Huang, Zaiyuan Wang, Xiaolong Chang · 10 juin 2026

  Recent years have witnessed the rapid evolution of AI agents toward handling increasingly complex, real-world tasks. However, existing benchmarks rarely evaluate whether agents can operate graphical user interfaces to complete long-horizon, high-value professional workflows across diverse domains. C…

• Visual-TCAV: Concept-based Attribution and Saliency Maps for Post-hoc Explainability in Image Classification

  Antonio De Santis, Riccardo Campi, Matteo Bianchi, Marco Brambilla · 10 juin 2026

  Convolutional Neural Networks (CNNs) have shown remarkable performance in image classification. However, interpreting their predictions is challenging due to the size and complexity of these models. State-of-the-art saliency methods generate local explanations highlighting the area in the input imag…

• One Lens, Many Worlds : A Capability-Typed Interface for World-Model Interpretability

  Bhavith Chandra Challagundla, Sanskar Pandey, Param Thakkar, Rishikesh Mallagundla, Yugandhar Reddy Gogireddy, Wenhao Lu, Hindol Roy Choudhury, Shravani Challagundla, Mohamed Deraz Nasr, Spursh Deshpande · 10 juin 2026

  World models are now built on substantially different computational substrates. Latent recurrent state-space models such as PlaNet and the Dreamer family compress observations into recurrent states; token-based models such as IRIS quantize observations into a learned codebook and predict autoregress…

• Does Normalization Choice Matter for Causal Large Time-Series Models?

  Samy-Melwan Vilhes (LMAC), Gilles Gasso (LMAC), Mokhtar Z Alaya (LMAC) · 10 juin 2026

  Large models for time-series forecasting have been emerged as a promising paradigm for training models on heterogeneous collections of signals. These models typically rely on causal autoregressive architectures, where each observation is sequentially predicted from past. In practice, real-world time…

• The Arbiter Agent: Continually Monitoring Multi-Agent Conversations to Detect Emergent Misalignment

  Filippo Tonini, Federico Torrielli, Anton Danholt Lautrup, Peter Schneider-Kamp, Mustafa Mert \c{C}elikok, Lukas Galke Poech · 10 juin 2026

  As AI systems built from multiple language-model agents become more common, they are increasingly used to make decisions together: discussing, negotiating, and acting on shared tasks. While individual agents may appear well-aligned when tested on their own, problems can arise from how they interact …

• Superficial Beliefs in LLM Decision-Making

  Gabriel Freedman, Francesca Toni · 10 juin 2026

  We ask whether large language models (LLMs) merely imitate rationales when choosing between two options, or whether their choices reflect a systematic underlying decision structure. Using synthetic binary decision settings in which models choose between profiles defined by graded attributes, we comp…

• Beyond Explaining Predictions: Logic-Based Explanations for Confidence in Machine Learning Models

  Vin\'icius Peixoto Chagas, Carlos Henrique Leit\~ao Cavalcante, Thiago Alves Rocha · 10 juin 2026

  Machine learning is increasingly used in critical domains, where both predictions and their associated confidence levels influence important decisions. To enhance transparency in such scenarios, it is important to understand why a model is confident or uncertain about its predictions. Recent logic-b…

• Trace2Policy: From Expert Behavior Traces to Self-Evolving Decision Agents

  Junli Zha, Jinbo Wang, Chao Zhou, Xiang Song · 10 juin 2026

  Decision rules that enterprise experts apply tacitly -- in auditing, compliance, and contract review -- can be systematically recovered and improved through iterative error analysis. We present \textbf{Trace2Policy}, whose core mechanism -- \textbf{EISR} (\textbf{E}rror-driven \textbf{I}terative \te…

• In Defense of Information Leakage in Concept-based Models

  Mateo Espinosa Zarlenga · 10 juin 2026

  Concept-based models (CMs), deep neural networks that ground their predictions on representations aligned with human-understandable concepts (e.g., "round", "stripes", etc.), have been shown to learn representations that leak concept-irrelevant information. As the traditional narrative goes, this le…

• Does Capability Transfer to Subjective Behavior -- and Would Our Instruments Tell Us? A Self-Evolving, Trust-by-Construction Evaluation Paradigm

  Yuming (Rapheal), Huang, Yao Liu, Pengjie Ding, Lei Wang, Junchen Wan · 10 juin 2026

  Benchmarking is mature where answers are verifiable -- math, code, reasoning -- but the fastest-growing uses of LLMs are subjective and human-facing: companionship, emotional support, counseling. There the default validity test, correlating a metric to human judgment, has no stable anchor: inter-rat…

• TD-Grokking: Learning from Zero-Reward Problems by Training-Time Decomposition

  Ningyuan Xi, Hao Xu, Hongsheng Xin, Ning Miao · 10 juin 2026

  Large language models (LLMs) have made remarkable progress in reasoning tasks, largely driven by post-training paradigms, especially reinforcement learning with verifiable rewards (RLVR). However, a critical bottleneck persists: RLVR fails on highly challenging zero-reward problems, where all sample…

• Generative Explainability for Next-Generation Networks: LLM-Augmented XAI with Mutual Feature Interactions

  Kiarash Rezaei, Omran Ayoub, Sebastian Troia, Francesco Lelli, Paolo Monti, Carlos Natalino · 10 juin 2026

  As artificial intelligence and machine learning (AI/ML) models become integral to network operations, their lack of transparency poses a significant barrier to operator trust. Existing explainable artificial intelligence (XAI) techniques often fail to bridge this gap for non-specialists, producing t…

• Can Global XAI Methods Reveal Injected Behaviours in LLMs? SHAP vs Rule Extraction vs RuleSHAP

  Francesco Sovrano · 9 juin 2026

  Large language models (LLMs) can amplify misinformation, undermining societal goals such as the UN SDGs. We study three documented drivers of misinformation (valence framing, information overload, and oversimplification) often shaped by default beliefs. Building on evidence that LLMs encode such def…

• Closure-Validated Circuit Discovery in Attention Heads: Co-activation Proposes, Ablation Disposes

  Yongzhong Xu · 9 juin 2026

  Interpretability increasingly treats groups of components, not individual units, as the basic object, and proposes to find them by clustering co-activation statistics. We ask whether such a cheap signal actually identifies an attention-head circuit. Adapting a sparse-autoencoder clustering recipe to…

• Position: Genomic Model Research Must Move Beyond Anecdotal Evaluation of Interpretability Methods

  Shasha Zhou, Mingyu Huang, Ke Li · 9 juin 2026

  Advances in machine learning and computational power have unlocked the predictive potential of the human genome, yet biologists now demand that these models also elucidate the underlying biological mechanisms. While interpretable machine learning (IML) techniques have been increasingly applied to br…

• Agentic Search for Counterfactual Recourse under Fixed LLM Budgets

  Yasuo Tabei · 9 juin 2026

  Counterfactual recourse aims to provide actionable feature changes that would alter an unfavorable decision made by a predictive model. In practice, affected individuals often benefit from multiple feasible alternatives rather than a single optimal explanation. A natural way to produce such alternat…

• Causal Agent Replay: Counterfactual Attribution for LLM-Agent Failures

  Jaineet Shah · 9 juin 2026

  When an LLM agent fails -- issues a refund it should not have, calls the wrong tool, leaks data -- existing tooling answers what happened (observability) or whether it passed (evaluation), but not which step caused the failure. The obvious heuristics are wrong: the step that executes the harmful act…

• When Are Neural Interaction Discoveries Real? Identifiability, Recoverability, and a Pre-Fit Diagnostic

  Valentina Kuskova, Dmitry Zaytsev, Michael Coppedge · 9 juin 2026

  When a neural time-series model reports that one variable modulates another's effect on a target, is the discovered interaction a property of the data or an artifact of model flexibility? We argue that this is fundamentally a question of identifiability, governed by the geometry of the observed inpu…

• Instrumented data for causal scientific machine learning

  Daniel N. Wilke · 9 juin 2026

  Scientific machine learning is limited less by model size than by the data it is trained on. Observational data records what happened but not why; template synthetic data has a known generating process but only for the simulator's template, not the case a user faces. We argue a third option is now o…

• Distilling LLM Reasoning into an Interpretable Policy Tree for Human-AI Collaboration

  Beiwen Zhang, Yongheng Liang, Guowei Zou, Haitao Wang, Hejun Wu · 9 juin 2026

  Constructing efficient and reliable policies to assist humans is indispensable for human-AI collaboration. Existing methods mainly follow two lines of work. Most prior work relies on multi-agent reinforcement learning (MARL) to learn black-box policies, which limits interpretability and raises safet…

• Self-Explainability in Self-Adaptive and Self-Organising Systems: Status and Research Directions

  Tom Beyer, Svea Wisy, Sven Tomforde · 9 juin 2026

  The growing complexity of self-adaptive and self-organising systems, fuelled by advances in Artificial Intelligence (AI), has made them increasingly difficult to understand and trust. While Explainable AI aims to provide insight into AI decision-making, a more advanced goal is for systems to explain…

• mllm-shap: A Shapley Value Explainability Platform for Text-Audio Multimodal Large Language Models

  Jakub Muszy\'nski, Pawe{\l} Pozorski, Maria Ganzha · 9 juin 2026

  We introduce mllm-shap, an open-source Python framework designed to extend Shapley Value (SV) explainability from text-only Large Language Models to Multimodal LLMs (MLLMs) processing joint text and audio inputs. While text-based attribution is well-studied, mllm-shap addresses three critical challe…

• How Reliable are Fairness Audits with Unreliable Data?

  Yash Vardhan Tomar · 9 juin 2026

  Fairness audits are a key component of responsible machine-learning deployment. Yet, the reliability of audit recommendations under incomplete protected-label access is still poorly understood. In this work, we focused on protected-label missingness in fairness mitigation audits. We introduced a see…

• Reasoning Arena: Trace Tournaments When Verifiable Rewards Fall Short

  Han Zhou, Adam X. Yang, Laurence Aitchison, Anna Korhonen, Albert Q. Jiang · 9 juin 2026

  Reinforcement learning with verifiable rewards (RLVR) has become a leading paradigm for improving the reasoning ability of large language models through outcome-based supervision. However, verifiable rewards frequently become uninformative at the group level: when all sampled traces of a given promp…

• SAEExplainer: Interpreting SAE Features with Activation-Guided Preference Optimization

  Jingyi He, Haiyan Zhao, Ruxue Shi, Yanguang Liu, Xin Wang, Fei Sun, Mengnan Du · 9 juin 2026

  Although Sparse Autoencoders (SAEs) have mitigated the opacity of large language models (LLMs) by decomposing dense representations into sparse features, explaining these features still remains a central challenge. Current explanation methods, however, typically operate within an open-loop paradigm,…