Explainable Artificial Intelligence (XAI)

Derniers papiers

• Don't Ignore the Tail: Decoupling top-K Probabilities for Efficient Language Model Distillation

  Sayantan Dasgupta, Trevor Cohn, Timothy Baldwin · 25 février 2026

  The core learning signal used in language model distillation is the standard Kullback-Leibler (KL) divergence between the student and teacher distributions. Traditional KL divergence tends to be dominated by the next tokens with the highest probabilities, i.e., the teacher's modes, thereby diminishi…

• Knowing the Unknown: Interpretable Open-World Object Detection via Concept Decomposition Model

  Xueqiang Lv, Shizhou Zhang, Yinghui Xing, Di Xu, Peng Wang, Yanning Zhang · 25 février 2026

  Open-world object detection (OWOD) requires incrementally detecting known categories while reliably identifying unknown objects. Existing methods primarily focus on improving unknown recall, yet overlook interpretability, often leading to known-unknown confusion and reduced prediction reliability. T…

• Counterfactual Simulation Training for Chain-of-Thought Faithfulness

  Peter Hase, Christopher Potts · 25 février 2026

  Inspecting Chain-of-Thought reasoning is among the most common means of understanding why an LLM produced its output. But well-known problems with CoT faithfulness severely limit what insights can be gained from this practice. In this paper, we introduce a training method called Counterfactual Simul…

• ContextPilot: Fast Long-Context Inference via Context Reuse

  Yinsicheng Jiang, Yeqi Huang, Liang Cheng, Cheng Deng, Xuan Sun, Luo Mai · 25 février 2026

  AI applications increasingly depend on long-context inference, where LLMs consume substantial context to support stronger reasoning. Common examples include retrieval-augmented generation, agent memory layers, and multi-agent orchestration. As input contexts get longer, prefill latency becomes the m…

• cc-Shapley: Measuring Multivariate Feature Importance Needs Causal Context

  J\"org Martin, Stefan Haufe · 25 février 2026

  Explainable artificial intelligence promises to yield insights into relevant features, thereby enabling humans to examine and scrutinize machine learning models or even facilitating scientific discovery. Considering the widespread technique of Shapley values, we find that purely data-driven operatio…

• Transcoder Adapters for Reasoning-Model Diffing

  Nathan Hu, Jake Ward, Thomas Icard, Christopher Potts · 25 février 2026

  While reasoning models are increasingly ubiquitous, the effects of reasoning training on a model's internal mechanisms remain poorly understood. In this work, we introduce transcoder adapters, a technique for learning an interpretable approximation of the difference in MLP computation before and aft…

• ConceptRM: The Quest to Mitigate Alert Fatigue through Consensus-Based Purity-Driven Data Cleaning for Reflection Modelling

  Yongda Yu, Lei Zhang, Xinxin Guo, Minghui Yu, Zhengqi Zhuang, Guoping Rong, Haifeng Shen, Zhengfeng Li, Boge Wang, Guoan Zhang, Bangyu Xiang, Xiaobin Xu · 25 février 2026

  In many applications involving intelligent agents, the overwhelming volume of alerts (mostly false) generated by the agents may desensitize users and cause them to overlook critical issues, leading to the so-called ''alert fatigue''. A common strategy is to train a reflection model as a filter to in…

• SELAUR: Self Evolving LLM Agent via Uncertainty-aware Rewards

  Dengjia Zhang, Xiaoou Liu, Lu Cheng, Yaqing Wang, Kenton Murray, Hua Wei · 25 février 2026

  Large language models (LLMs) are increasingly deployed as multi-step decision-making agents, where effective reward design is essential for guiding learning. Although recent work explores various forms of reward shaping and step-level credit assignment, a key signal remains largely overlooked: the i…

• Right to History: A Sovereignty Kernel for Verifiable AI Agent Execution

  Jing Zhang · 25 février 2026

  AI agents increasingly act on behalf of humans, yet no existing system provides a tamper-evident, independently verifiable record of what they did. As regulations such as the EU AI Act begin mandating automatic logging for high-risk AI systems, this gap carries concrete consequences -- especially fo…

• "Are You Sure?": An Empirical Study of Human Perception Vulnerability in LLM-Driven Agentic Systems

  Xinfeng Li, Shenyu Dai, Kelong Zheng, Yue Xiao, Gelei Deng, Wei Dong, Xiaofeng Wang · 25 février 2026

  Large language model (LLM) agents are rapidly becoming trusted copilots in high-stakes domains like software development and healthcare. However, this deepening trust introduces a novel attack surface: Agent-Mediated Deception (AMD), where compromised agents are weaponized against their human users.…

• Watson & Holmes: A Naturalistic Benchmark for Comparing Human and LLM Reasoning

  Thatchawin Leelawat, Lewis D Griffin · 24 février 2026

  Existing benchmarks for AI reasoning provide limited insight into how closely these capabilities resemble human reasoning in naturalistic contexts. We present an adaptation of the Watson & Holmes detective tabletop game as a new benchmark designed to evaluate reasoning performance using incrementall…

• TRUE: A Trustworthy Unified Explanation Framework for Large Language Model Reasoning

  Yujiao Yang · 24 février 2026

  Large language models (LLMs) have demonstrated strong capabilities in complex reasoning tasks, yet their decision-making processes remain difficult to interpret. Existing explanation methods often lack trustworthy structural insight and are limited to single-instance analysis, failing to reveal reas…

• Defining Explainable AI for Requirements Analysis

  Raymond Sheh, Isaac Monteath · 24 février 2026

  Explainable Artificial Intelligence (XAI) has become popular in the last few years. The Artificial Intelligence (AI) community in general, and the Machine Learning (ML) community in particular, is coming to the realisation that in many applications, for AI to be trusted, it must not only demonstrate…

• Rules or Weights? Comparing User Understanding of Explainable AI Techniques with the Cognitive XAI-Adaptive Model

  Louth Bin Rawshan, Zhuoyu Wang, Brian Y Lim · 24 février 2026

  Rules and Weights are popular XAI techniques for explaining AI decisions. Yet, it remains unclear how to choose between them, lacking a cognitive framework to compare their interpretability. In an elicitation user study on forward and counterfactual decision tasks, we identified 7 reasoning strategi…

• The Story is Not the Science: Execution-Grounded Evaluation of Mechanistic Interpretability Research

  Xiaoyan Bai, Alexander Baumgartner, Haojia Sun, Ari Holtzman, Chenhao Tan · 24 février 2026

  Reproducibility crises across sciences highlight the limitations of the paper-centric review system in assessing the rigor and reproducibility of research. AI agents that autonomously design and generate large volumes of research outputs exacerbate these challenges. In this work, we address the grow…

• ComplLLM: Fine-tuning LLMs to Discover Complementary Signals for Decision-making

  Ziyang Guo, Yifan Wu, Jason Hartline, Kenneth Holstein, Jessica Hullman · 24 février 2026

  Multi-agent decision pipelines can outperform single agent workflows when complementarity holds, i.e., different agents bring unique information to the table to inform a final decision. We propose ComplLLM, a post-training framework based on decision theory that fine-tunes a decision-assistant LLM u…

• Bayesian Lottery Ticket Hypothesis

  Nicholas Kuhn, Arvid Weyrauch, Lars Heyen, Achim Streit, Markus G\"otz, Charlotte Debus · 24 février 2026

  Bayesian neural networks (BNNs) are a useful tool for uncertainty quantification, but require substantially more computational resources than conventional neural networks. For non-Bayesian networks, the Lottery Ticket Hypothesis (LTH) posits the existence of sparse subnetworks that can train to the …

• Winsor-CAM: Human-Tunable Visual Explanations from Deep Networks via Layer-Wise Winsorization

  Casey Wall, Longwei Wang, Rodrigue Rizk, KC Santosh · 24 février 2026

  Interpreting Convolutional Neural Networks (CNNs) is critical for safety-sensitive applications such as healthcare and autonomous systems. Popular visual explanation methods like Grad-CAM use a single convolutional layer, potentially missing multi-scale cues and producing unstable saliency maps. We …

• Understanding Empirical Unlearning with Combinatorial Interpretability

  Shingo Kodama, Niv Cohen, Micah Adler, Nir Shavit · 24 février 2026

  While many recent methods aim to unlearn or remove knowledge from pretrained models, seemingly erased knowledge often persists and can be recovered in various ways. Because large foundation models are far from interpretable, understanding whether and how such knowledge persists remains a significant…

• IAPO: Information-Aware Policy Optimization for Token-Efficient Reasoning

  Yinhan He, Yaochen Zhu, Mingjia Shi, Wendy Zheng, Lin Su, Xiaoqing Wang, Qi Guo, Jundong Li · 24 février 2026

  Large language models increasingly rely on long chains of thought to improve accuracy, yet such gains come with substantial inference-time costs. We revisit token-efficient post-training and argue that existing sequence-level reward-shaping methods offer limited control over how reasoning effort is …

• Detecting Cybersecurity Threats by Integrating Explainable AI with SHAP Interpretability and Strategic Data Sampling

  Norrakith Srisumrith, Sunantha Sodsee · 24 février 2026

  The critical need for transparent and trustworthy machine learning in cybersecurity operations drives the development of this integrated Explainable AI (XAI) framework. Our methodology addresses three fundamental challenges in deploying AI for threat detection: handling massive datasets through Stra…

• Could Large Language Models work as Post-hoc Explainability Tools in Credit Risk Models?

  Wenxi Geng, Dingyuan Liu, Liya Li, Yiqing Wang · 24 février 2026

  Post-hoc explainability is central to credit risk model governance, yet widely used tools such as coefficient-based attributions and SHapley Additive exPlanations (SHAP) often produce numerical outputs that are difficult to communicate to non-technical stakeholders. This paper investigates whether l…

• Adaptive Collaboration of Arena-Based Argumentative LLMs for Explainable and Contestable Legal Reasoning

  Hoang-Loc Cao, Phuc Ho, Truong Thanh Hung Nguyen, Phuc Truong Loc Nguyen, Dinh Thien Loc Nguyen, Hung Cao · 24 février 2026

  Legal reasoning requires not only high accuracy but also the ability to justify decisions through verifiable and contestable arguments. However, existing Large Language Model (LLM) approaches, such as Chain-of-Thought (CoT) and Retrieval-Augmented Generation (RAG), often produce unstructured explana…

• The Confusion is Real: GRAPHIC - A Network Science Approach to Confusion Matrices in Deep Learning

  Johanna S. Fr\"ohlich, Bastian Heinlein, Jan U. Claar, Hans Rosenberger, Vasileios Belagiannis, Ralf R. M\"uller · 24 février 2026

  Explainable artificial intelligence has emerged as a promising field of research to address reliability concerns in artificial intelligence. Despite significant progress in explainable artificial intelligence, few methods provide a systematic way to visualize and understand how classes are confused …

• EconCausal: A Context-Aware Causal Reasoning Benchmark for Large Language Models in Social Science

  Donggyu Lee, Hyeok Yun, Meeyoung Cha, Sungwon Park, Sangyoon Park, Jihee Kim · 24 février 2026

  Socio-economic causal effects depend heavily on their specific institutional and environmental context. A single intervention can produce opposite results depending on regulatory or market factors, contexts that are often complex and only partially observed. This poses a significant challenge for la…