Explainable Artificial Intelligence (XAI)

Derniers papiers

• One Permutation Is All You Need: Fast, Reliable Variable Importance and Model Stress-Testing

  Albert Dorador · 17 décembre 2025

  Reliable estimation of feature contributions in machine learning models is essential for trust, transparency and regulatory compliance, especially when models are proprietary or otherwise operate as black boxes. While permutation-based methods are a standard tool for this task, classical implementat…

• Non-Resolution Reasoning (NRR): A Computational Framework for Contextual Identity and Ambiguity Preservation

  Kei Saito · 17 décembre 2025

  Current artificial intelligence systems, despite remarkable capabilities in text generation and pattern recognition, exhibit a fundamental architectural limitation: they resolve ambiguity prematurely. This premature semantic collapse -- the tendency to collapse multiple valid interpretations into a …

• Ladder Up, Memory Down: Low-Cost Fine-Tuning With Side Nets

  Estelle Zheng (LORIA, ALE), Nathan Cerisara (LORIA), S\'ebastien Warichet (ALE), Emmanuel Helbert (ALE), Christophe Cerisara (SYNALP, LORIA) · 17 décembre 2025

  Fine-tuning large language models (LLMs) is often limited by the memory available on commodity GPUs. Parameter-efficient fine-tuning (PEFT) methods such as QLoRA reduce the number of trainable parameters, yet still incur high memory usage induced by the backward pass in the full model. We revisit La…

• ValuePilot: A Two-Phase Framework for Value-Driven Decision-Making

  Yitong Luo, Ziang Chen, Hou Hei Lam, Jiayu zhan, Junqi Wang, Zhenliang Zhang, Xue Feng · 17 décembre 2025

  Personalized decision-making is essential for human-AI interaction, enabling AI agents to act in alignment with individual users' value preferences. As AI systems expand into real-world applications, adapting to personalized values beyond task completion or collective alignment has become a critical…

• Explainable reinforcement learning from human feedback to improve alignment

  Shicheng Liu, Siyuan Xu, Wenjie Qiu, Hangfan Zhang, Minghui Zhu · 17 décembre 2025

  A common and effective strategy for humans to improve an unsatisfactory outcome in daily life is to find a cause of this outcome and correct the cause. In this paper, we investigate whether this human improvement strategy can be applied to improving reinforcement learning from human feedback (RLHF) …

• Counterfactual Explanations for Time Series Should be Human-Centered and Temporally Coherent in Interventions

  Emmanuel C. Chukwu, Rianne M. Schouten, Monique Tabak, Mykola Pechenizkiy · 17 décembre 2025

  Counterfactual explanations are increasingly proposed as interpretable mechanisms to achieve algorithmic recourse. However, current counterfactual techniques for time series classification are predominantly designed with static data assumptions and focus on generating minimal input perturbations to …

• Enhancing Interpretability for Vision Models via Shapley Value Optimization

  Kanglong Fan, Yunqiao Yang, Chen Ma · 17 décembre 2025

  Deep neural networks have demonstrated remarkable performance across various domains, yet their decision-making processes remain opaque. Although many explanation methods are dedicated to bringing the obscurity of DNNs to light, they exhibit significant limitations: post-hoc explanation methods ofte…

• Inverse Scaling in Test-Time Compute

  Aryo Pradipta Gema, Alexander H\"agele, Runjin Chen, Andy Arditi, Jacob Goldman-Wetzler, Kit Fraser-Taliente, Henry Sleight, Linda Petrini, Julian Michael, Beatrice Alex, Pasquale Minervini, Yanda Chen, Joe Benton, Ethan Perez · 17 décembre 2025

  We construct evaluation tasks where extending the reasoning length of Large Reasoning Models (LRMs) deteriorates performance, exhibiting an inverse scaling relationship between test-time compute and accuracy. Our evaluation tasks span four categories: simple counting tasks with distractors, regressi…

• Optimized Architectures for Kolmogorov-Arnold Networks

  James Bagrow, Josh Bongard · 16 décembre 2025

  Efforts to improve Kolmogorov-Arnold networks (KANs) with architectural enhancements have been stymied by the complexity those enhancements bring, undermining the interpretability that makes KANs attractive in the first place. Here we study overprovisioned architectures combined with sparsification …

• AgentSHAP: Interpreting LLM Agent Tool Importance with Monte Carlo Shapley Value Estimation

  Miriam Horovicz · 16 décembre 2025

  LLM agents that use external tools can solve complex tasks, but understanding which tools actually contributed to a response remains a blind spot. No existing XAI methods address tool-level explanations. We introduce AgentSHAP, the first framework for explaining tool importance in LLM agents. AgentS…

• On the Accuracy of Newton Step and Influence Function Data Attributions

  Ittai Rubinstein, Samuel B. Hopkins · 16 décembre 2025

  Data attribution aims to explain model predictions by estimating how they would change if certain training points were removed, and is used in a wide range of applications, from interpretability and credit assignment to unlearning and privacy. Even in the relatively simple case of linear regressio…

• Causal inference and model explainability tools for retail

  Pranav Gupta, Nithin Surendran · 16 décembre 2025

  Most major retailers today have multiple divisions focused on various aspects, such as marketing, supply chain, online customer experience, store customer experience, employee productivity, and vendor fulfillment. They also regularly collect data corresponding to all these aspects as dashboards and …

• CR3G: Causal Reasoning for Patient-Centric Explanations in Radiology Report Generation

  Satyam Kumar · 16 décembre 2025

  Automatic chest X-ray report generation is an important area of research aimed at improving diagnostic accuracy and helping doctors make faster decisions. Current AI models are good at finding correlations (or patterns) in medical images. Still, they often struggle to understand the deeper cause-and…

• AI-Assisted Game Management Decisions: A Fuzzy Logic Approach to Real-Time Soccer Substitutions

  Pedro Passos · 16 décembre 2025

  In elite soccer, substitution decisions entail significant financial and sporting consequences yet remain heavily reliant on intuition or predictive models that merely mimic historical biases. This paper introduces a Fuzzy Logic based Decision Support System (DSS) designed for real time, prescriptiv…

• Data-Driven Global Sensitivity Analysis for Engineering Design Based on Individual Conditional Expectations

  Pramudita Satria Palar, Paul Saves, Rommel G. Regis, Koji Shimoyama, Shigeru Obayashi, Nicolas Verstaevel, Joseph Morlier · 16 décembre 2025

  Explainable machine learning techniques have gained increasing attention in engineering applications, especially in aerospace design and analysis, where understanding how input variables influence data-driven models is essential. Partial Dependence Plots (PDPs) are widely used for interpreting black…

• Resting Neurons, Active Insights: Improving Input Sparsification for Large Language Models

  Haotian Xu, Tian Gao, Tsui-Wei Weng, Tengfei Ma · 16 décembre 2025

  Large Language Models (LLMs) achieve state-of-the-art performance across a wide range of applications, but their massive scale poses significant challenges for both efficiency and interpretability. Structural pruning, which reduces model size by removing redundant computational units such as neurons…

• Explainable Artificial Intelligence for Economic Time Series: A Comprehensive Review and a Systematic Taxonomy of Methods and Concepts

  Agust\'in Garc\'ia-Garc\'ia, Pablo Hidalgo, Julio E. Sandubete · 16 décembre 2025

  Explainable Artificial Intelligence (XAI) is increasingly required in computational economics, where machine-learning forecasters can outperform classical econometric models but remain difficult to audit and use for policy. This survey reviews and organizes the growing literature on XAI for economic…

• State over Tokens: Characterizing the Role of Reasoning Tokens

  Mosh Levy, Zohar Elyoseph, Shauli Ravfogel, Yoav Goldberg · 16 décembre 2025

  Large Language Models (LLMs) can generate reasoning tokens before their final answer to boost performance on complex tasks. While these sequences seem like human thought processes, empirical evidence reveals that they are not a faithful explanation of the model's actual reasoning process. To address…

• Non-Resolution Reasoning: A Framework for Preserving Semantic Ambiguity in Language Models

  Kei Saito · 16 décembre 2025

  Premature semantic collapse -- the forced early commitment to a single meaning -- remains a core architectural limitation of current language models. Softmax-driven competition and greedy decoding cause models to discard valid interpretations before sufficient context is available, resulting in brit…

• Soft Decision Tree classifier: explainable and extendable PyTorch implementation

  Reuben R Shamir · 16 décembre 2025

  We implemented a Soft Decision Tree (SDT) and a Short-term Memory Soft Decision Tree (SM-SDT) using PyTorch. The methods were extensively tested on simulated and clinical datasets. The SDT was visualized to demonstrate the potential for its explainability. SDT, SM-SDT, and XGBoost demonstrated simil…

• Think, Speak, Decide: Language-Augmented Multi-Agent Reinforcement Learning for Economic Decision-Making

  Heyang Ma, Qirui Mi, Qipeng Yang, Zijun Fan, Bo Li, Haifeng Zhang · 16 décembre 2025

  Economic decision-making depends not only on structured signals such as prices and taxes, but also on unstructured language, including peer dialogue and media narratives. While multi-agent reinforcement learning (MARL) has shown promise in optimizing economic decisions, it struggles with the semanti…

• Value-Aware Multiagent Systems

  Nardine Osman · 16 décembre 2025

  This paper introduces the concept of value awareness in AI, which goes beyond the traditional value-alignment problem. Our definition of value awareness presents us with a concise and simplified roadmap for engineering value-aware AI. The roadmap is structured around three core pillars: (1) learning…

• FROC: A Unified Framework with Risk-Optimized Control for Machine Unlearning in LLMs

  Si Qi Goh, Yongsen Zheng, Ziyao Liu, Sami Hormi, Kwok-Yan Lam · 16 décembre 2025

  Machine unlearning (MU) seeks to eliminate the influence of specific training examples from deployed models. As large language models (LLMs) become widely used, managing risks arising from insufficient forgetting or utility loss is increasingly crucial. Current MU techniques lack effective mechanism…

• A neuro-symbolic framework for accountability in public-sector AI

  Allen Daniel Sunny · 16 décembre 2025

  Automated eligibility systems increasingly determine access to essential public benefits, but the explanations they generate often fail to reflect the legal rules that authorize those decisions. This thesis develops a legally grounded explainability framework that links system-generated decision jus…

• XNNTab -- Interpretable Neural Networks for Tabular Data using Sparse Autoencoders

  Khawla Elhadri, J\"org Schl\"otterer, Christin Seifert · 16 décembre 2025

  In data-driven applications relying on tabular data, where interpretability is key, machine learning models such as decision trees and linear regression are applied. Although neural networks can provide higher predictive performance, they are not used because of their blackbox nature. In this work, …