Explainable Artificial Intelligence (XAI)

Derniers papiers

• Calibrated Selective Fact-Checking via Evidence Chain Evaluation

  Dekun Yang · 22 juillet 2026

  Large language models (LLMs) can achieve strong fact-checking accuracy, yet forced binary decisions conceal a critical reliability problem: systems may issue confident verdicts even when supporting evidence is weak, sparse, or internally inconsistent. We address this issue through Evidence Chain Eva…

• Reliability Scales Inversely: Bigger Models Compound Mistakes Faster via a Hidden Auto-Regressive Risk Regime

  Kushal Chakrabarti · 22 juillet 2026

  As language models scale, answers start truer but degrade faster: scaling buys capability but erodes reliability. The knowledge-gap account - more data, retrieval, or scale - misses an auto-regressive risk residual that scale sharpens: the model commits to a low-probability token, conditions on it a…

• Fishing Out Free Riders: Shapley-Based Reward Attribution for Parallel Reasoning via Reinforcement Learning

  Wentao Zhang, Haoyu Zhang, Xinke Jiang, Yuxuan Cheng, Yuhan Pan, Miao Li, Zhipeng Qiao, Tao Feng, Zhen Tao, Dengji Zhao · 22 juillet 2026

  Large Language Models (LLMs) excel at multi-step reasoning, yet current parallel reasoning approaches often fail to distinguish the contributions of individual reasoning paths. Many paths may be redundant, misleading, or even detrimental, but outcome-level rewards assign uniform reward, leading to a…

• Weak-to-Strong Learning in Decision Making

  Jingwei Ji, Renyuan Xu · 22 juillet 2026

  Many operational decisions rely on predictive models that estimate uncertain outcomes conditional on observable contexts. Training such models, however, often faces a fundamental data asymmetry: labeled outcomes are scarce or costly to obtain, while contextual covariates are abundant. Motivated by t…

• ConceptCF: Concept-based Counterfactuals for the Explainability of Time Series

  Annemarie Jutte, Faizan Ahmed, Jeroen Linssen, Maurice van Keulen · 22 juillet 2026

  This paper proposes ConceptCF, a method for counterfactual generation that operates on human-interpretable concepts. In high-stakes domains such as healthcare and predictive maintenance, artificial intelligence models can increase efficiency and safety. Explainability is key to ensure these models r…

• Spectral Higher-Order Neural Networks Have Sharp Expressivity Bounds

  Gianluca Peri, Diego Febbe, Duccio Fanelli · 22 juillet 2026

  Neural hypergraphs are a natural generalization of neural networks, the reference models in modern machine learning. Yet, their deployment has proven demanding: the number of weighted hyperedges required leads to an intractable parameter explosion. However, a novel parametrization that leverages spe…

• Using LLMs for Explainable, Data-Driven Insight Generation from Time Series

  Ria Mundhra, Gustavo Sato dos Santos, Michael Benedikt · 22 juillet 2026

  Time series forecasts are widely used in decision-critical domains, where they are rarely consumed without accompanying explanations. Producing such explanations is usually a manual and costly process, and attempts to automate it using large language models often suffer from hallucination when appli…

• Human Grounded Evaluation of Large Language Models for Optical Network Automation

  Kiarash Rezaei, Omran Ayoub, Paolo Monti, Carlos Natalino · 21 juillet 2026

  Large language models (LLMs) are increasingly adopted for network automation, yet their output quality and inference cost can vary substantially across LLM families. We present HuGLEN, a stepwise evaluation pipeline that uses an LLM-as-a-judge together with a small set of expert ratings to enable sc…

• FST.ai 2.5: Explainable and Uncertainty-Aware AI for Olympic and Para-Taekwondo Decision Support, Athlete Digital Twins, and Federation-Scale Analytics

  Keivan Shariatmadar, Ahmad Osman, Ramin Rey · 21 juillet 2026

  The rapid digitalisation of elite sport has created new opportunities for integrating artificial intelligence (AI), performance analytics, and decision-support systems into athlete development and competition management. However, existing solutions remain fragmented, typically addressing isolated ta…

• Can Interpretation Predict Behavior on Unseen Data?

  Victoria R. Li, Jenny Kaufmann, Tian Qin, Martin Wattenberg, David Alvarez-Melis, Naomi Saphra · 21 juillet 2026

  Interpretability research often predicts model responses to targeted mechanistic interventions. But can we predict responses to unseen input data? We propose and demonstrate this alternate objective by using model internals to predict their out-of-distribution (OOD) behavior. We train hundreds of Tr…

• Expected Free Energy as Belief-Dependent Utility for rho-POMDPs

  Patrick Cooper, Alvaro Velasquez · 21 juillet 2026

  An agent acting under partial observability must decide when to gather information and which observations are worth their cost. Standard POMDPs value information only through its eventual effect on reward. The $\rho$-POMDP framework instead rewards uncertainty reduction directly, through a belief-de…

• A Method for Learning Value Systems in Generative AI

  Andr\'es Holgado-S\'anchez, Holger Billhardt, Sascha Ossowski · 21 juillet 2026

  Value-aware AI systems require explicit computational representations of human values (groundings) and their aggregation into value systems in order to align their decisions with ours. As such representations are difficult to elicit, value learning seeks to infer them by observing human behaviour. T…

• RECON: Benchmarking Agent Memory for Compositional Reasoning over Long Contexts

  Mihir Shriniwas Arya · 21 juillet 2026

  Large language models and LLM-based agents are widely used as personal chat assistants, enterprise copilots, and autonomous workflow agents. In all these applications, memory (the ability to retain, access, and reason over information accumulated over long contexts and multiple interactions) plays a…

• A Geometric Perspective on Stabilizing Value Conflict Resolution

  Saket Reddy, Andy Liu · 21 juillet 2026

  Large Language Models (LLMs) often struggle to navigate value conflicts when trained with the compressed scalar rewards of Reinforcement Learning from Human Feedback (RLHF). To address this challenge, we investigate how chain-of-thought (CoT) reasoning can help improve performance in this domain. Ge…

• Can Transformers Really Do It All? On the Compatibility of Inductive Biases Across Tasks

  Damien Teney, Liangze Jiang, Hemanth Saratchandran, Simon Lucey · 21 juillet 2026

  Transformers are remarkably versatile and their design is largely consistent across a variety of applications. But are they optimal for any given task or dataset? The answer may be key for pushing AI beyond merely scaling current designs. *Method.* We present a method to optimize a transformer arc…

• Value-Monotonicity Matters: A Concordance Loss for Deep Survival Prediction

  Meixu Chen, Kai Wang, Jing Wang · 21 juillet 2026

  Deep survival models are evaluated almost exclusively by the concordance index (C-index), yet they are commonly trained using likelihood objectives such as the Cox partial likelihood, discrete-time negative log-likelihood, and DeepHit likelihood. This mismatch is usually considered acceptable becaus…

• Breaking the Block: Preserving Data Continuity to Train Superior SAEs for Instruct Models

  Jiaming Li, Haoran Ye, Yukun Chen, Xinyue Li, Lei Zhang, Hamid Alinejad-Rokny, Jimmy Chih-Hsien Peng, Min Yang · 21 juillet 2026

  Sparse Autoencoders (SAEs) are a cornerstone of mechanistic interpretability. Existing training methods inherit the Block Training paradigm from LLM pre-training, which introduces destructive gradient noise in instruct models due to attention leakage from unrelated contexts. Using GSNR analysis, we …

• ConceptTree: Bringing Semantic Transparency to Black-Box Decision Making for Robotic Manipulation

  Yongyan Wen, Feifan Liu, Jinyi Chen, Bo An, Peng Liu, Siyuan Li · 21 juillet 2026

  Establishing interpretable decision-making processes in long-horizon robotic manipulation is critical for enabling reliable human oversight and intervention. However, existing approaches to robotic manipulation largely treat skill selection as opaque mappings from observations to actions, offering l…

• Counterfactual Shapley Credit Assignment

  Mingxuan Li, Kaizhan-Lee, Elias Bareinboim · 21 juillet 2026

  The Credit Assignment Problem (CAP) is fundamental to developing efficient and explainable Reinforcement Learning (RL) agents. Existing frameworks, whether relying on temporal contiguity or hindsight-conditioned reward reweighting, frequently fail to attribute properly between an agent's policy (ski…

• The Failures of Marginal Influence-Based Attribution Methods for Global Time Series Explanations

  Amadeo Tunyi · 21 juillet 2026

  Explainability methods for time series models predominantly produce flat attribution scores: they quantify the direct influence of a feature at a timestamp by a scalar. We prove that the dominant failure mode of such methods is not the scalar format itself but a fundamental computational mismatch: e…

• What does a Bayes-filtered transformer believe? A predictive Monte Carlo approach

  Afiq Abdillah Effiezal Aswadi, Haotong Ma, Susan Wei · 21 juillet 2026

  A Bayes-filtered transformer (BFT) is a transformer trained on sequences that are generated in two steps: first a latent task is drawn from a prior, then observations are drawn conditional on that task. Trained under autoregressive log loss, the BFT's next-token prediction, in the idealized limit, i…

• Explaining and Tuning Transformer-based LLMs in Arithmetic Tasks with Human Strategies

  Luyu Qiu, Jianing Li, Hwanhee Kim, Xiaoyong Wei, Yueyuan Zheng, Janet Hsiao, Lei Chen · 21 juillet 2026

  Transformer-based large language models (LLMs) continue to achieve state-of-the-art performance across various natural language processing tasks. However, their subpar performance on seemingly elementary problems, such as basic arithmetic, raises concerns about model reliability, safety, and ethical…

• K-IPO: Kendall-constrained Importance Preserving Oversampling for Imbalanced Tabular Data

  Marios Tyrovolas, Argiris Sofotasios, Dimitris Metaxakis, Georgios Mermigkis, George Georgoulas, Panagiotis Hadjidoukas, Chrysostomos Stylios · 21 juillet 2026

  Oversampling is widely used to address class imbalance in tabular classification, but existing methods can distort the feature importance ranking underlying model explanations. Although recent studies have quantified this distortion by comparing real and synthetic data, none have actively sought to …

• BACON: Budgeted Human Calibration for Modeling and Evaluation with Multiple AI Judges

  Lei Shi, Anlan Zhang, Rita Lyu, Zhengmian Hu, Tong Yu, David Arbour, Avi Feller, Saayan Mitra, Ritwik Sinha · 21 juillet 2026

  AI judges offer a scalable, low-cost alternative to human evaluation, but their outputs can be biased relative to human preferences and highly item-dependent, varying across judges, tasks, and domains. When uncalibrated AI evaluations are used for model ranking, item scoring, or population-level qua…

• Position: Explanation Stability Is a Property of the Model Method Pair, Not the Model

  Kabilan Elangovan, Daniel Ting · 21 juillet 2026

  This position paper argues that claims about explanation stability are scientifically invalid without cross method validation. Just as statistical significance requires the test statistic to be specified, stability should either be evaluated across multiple attribution paradigms or explicitly scoped…