arXiv:2511.09332v1 Announce Type: new 
Abstract: The proliferation of complex, black-box AI models has intensified the need for techniques that can explain their decisions. Feature attribution methods have become a popular solution for providing post-hoc explanations, yet the field has historically lacked a formal problem definition. This paper addresses this gap by introducing a formal definition for the problem of feature attribution, which stipulates that explanations be supported by an underlying probability distribution represented by the given dataset. Our analysis reveals that many existing model-agnostic methods fail to meet this criterion, while even those that do often possess other limitations. To overcome these challenges, we propose Distributional Feature Attribution eXplanations (DFAX), a novel, model-agnostic method for feature attribution. DFAX is the first feature attribution method to explain classifier predictions directly based on the data distribution. We show through extensive experiments that DFAX is more effective and efficient than state-of-the-art baselines.

تقدم ورقة جديدة تفسيرات توزيع السمات (DFAX)، وهي طريقة غير مرتبطة بنموذج لشرح توقعات مصنّفات الذكاء الاصطناعي بناءً على توزيع البيانات. يعالج هذا النقص في التعريفات الرسمية في نسبة السمات، مشيرًا إلى أن العديد من الطرق الحالية لا تلبي المعايير اللازمة. يُظهر DFAX فعاليته وكفاءته مقارنةً بالأساليب الحالية، مما يمثل تقدمًا كبيرًا في قابلية تفسير الذكاء الاصطناعي.

Un nuevo artículo presenta las Explicaciones de Atribución de Características Distribucionales (DFAX), un método agnóstico al modelo para explicar las predicciones de clasificadores de IA basado en la distribución de datos. Esto aborda la falta de definiciones formales en la atribución de características, destacando que muchos métodos existentes no cumplen con los criterios necesarios. DFAX se muestra más efectivo y eficiente que los métodos actuales, marcando un avance significativo en la interpretabilidad de la IA.

Un nouvel article présente les Explications d'Attribution de Caractéristiques Distributionnelles (DFAX), une méthode agnostique au modèle pour expliquer les prédictions des classificateurs d'IA basée sur la distribution des données. Cela répond au manque de définitions formelles dans l'attribution des caractéristiques, soulignant que de nombreuses méthodes existantes ne répondent pas aux critères nécessaires. DFAX s'avère plus efficace et efficient que les méthodes actuelles, marquant une avancée significative dans l'interprétabilité de l'IA.

A new paper introduces Distributional Feature Attribution eXplanations (DFAX), a model-agnostic method for explaining AI classifier predictions based on data distribution. This addresses the lack of formal definitions in feature attribution, highlighting that many existing methods do not meet necessary criteria. DFAX is shown to be more effective and efficient than current state-of-the-art methods, marking a significant advancement in AI interpretability.

Distribution-Based Feature Attribution for Explaining the Predictions of Any Classifier

arXiv:1906.07172v5 Announce Type: replace-cross 
Abstract: Equivariant neural networks are a class of neural networks designed to preserve symmetries inherent in the data. In this paper, we introduce a general method for modifying a neural network to enforce equivariance, a process we refer to as equivarification. We further show that group convolutional neural networks (G-CNNs) arise as a special case of our framework.

Equivariant neural networks and equivarification

arXiv:2311.13745v4 Announce Type: replace-cross 
Abstract: Diffusion models have become the most popular approach to deep generative modeling of images, largely due to their empirical performance and reliability. From a theoretical standpoint, a number of recent works have studied the iteration complexity of sampling, assuming access to an accurate diffusion model. In this work, we focus on understanding the sample complexity of training such a model; how many samples are needed to learn an accurate diffusion model using a sufficiently expressive neural network? Prior work showed bounds polynomial in the dimension, desired Total Variation error, and Wasserstein error. We show an exponential improvement in the dependence on Wasserstein error and depth, along with improved dependencies on other relevant parameters.

Improved Sample Complexity Bounds for Diffusion Model Training

arXiv:2410.22995v2 Announce Type: replace-cross 
Abstract: A hallmark of advanced artificial intelligence is the capacity to progress from passive visual perception to the strategic modification of visual information to facilitate complex reasoning. This advanced capability, however, remains critically underdeveloped in current Large Multi-modal Models (LMMs). The deficiency is often masked by evaluation metrics that prioritize final-answer accuracy, creating an illusion of competence where genuine reasoning is absent. Using the domain of geometric problem-solving as a precise instrument, we probe this issue through tasks that require constructing visual aids. To this end, we introduce \textbf{VisAidMath}, a challenging benchmark, and our novel Three-Layered Funnel Evaluation Framework. This framework moves beyond simple accuracy (ACCU) to scrutinize the generation of valid visual aids (PVA) and the soundness of subsequent reasoning steps (SPRS). Our extensive experiments on state-of-the-art models, including Doubao-Seed-1.6 and o4, reveal a profound ``Reasoning Illusion''. We observe that high surface-level accuracy conceals a catastrophic failure in the models' ability to produce valid visual aids or to reason from them. Our findings expose a fundamental schism between visual perception and logical deduction in modern LMMs. We host an evaluation platform at CodaBench for testing publicly. Homepage: https://nlp2ct.github.io/VisAidMathHomepage/ Evaluation: https://www.codabench.org/competitions/7634/

VisAidMath: Benchmarking Visual-Aided Mathematical Reasoning

arXiv:2508.01249v3 Announce Type: replace-cross 
Abstract: Large Language Model (LLM) agents offer a powerful new paradigm for solving various problems by combining natural language reasoning with the execution of external tools. However, their dynamic and non-transparent behavior introduces critical security risks, particularly in the presence of prompt injection attacks. In this work, we propose a novel insight that treats the agent runtime traces as structured programs with analyzable semantics. Thus, we present AgentArmor, a program analysis framework that converts agent traces into graph intermediate representation-based structured program dependency representations (e.g., CFG, DFG, and PDG) and enforces security policies via a type system. AgentArmor consists of three key components: (1) a graph constructor that reconstructs the agent's runtime traces as graph-based intermediate representations with control and data flow described within; (2) a property registry that attaches security-relevant metadata of interacted tools \& data, and (3) a type system that performs static inference and checking over the intermediate representation. By representing agent behavior as structured programs, AgentArmor enables program analysis for sensitive data flow, trust boundaries, and policy violations. We evaluate AgentArmor on the AgentDojo benchmark, the results show that AgentArmor can reduce the ASR to 3\%, with the utility drop only 1\%.

AgentArmor: Enforcing Program Analysis on Agent Runtime Trace to Defend Against Prompt Injection

arXiv:2511.14691v1 Announce Type: cross 
Abstract: Attention is the brain's ability to selectively focus on a few specific aspects while ignoring irrelevant ones. This biological principle inspired the attention mechanism in modern Transformers. Transformers now underpin large language models (LLMs) such as GPT, but at the cost of massive training and inference energy, leading to a large carbon footprint. While brain attention emerges from neural circuits, Transformer attention relies on dot-product similarity to weight elements in the input sequence. Neuromorphic computing, especially spiking neural networks (SNNs), offers a brain-inspired path to energy-efficient intelligence. Despite recent work on attention-based spiking Transformers, the core attention layer remains non-neuromorphic. Current spiking attention (i) relies on dot-product or element-wise similarity suited to floating-point operations, not event-driven spikes; (ii) keeps attention matrices that suffer from the von Neumann bottleneck, limiting in-memory computing; and (iii) still diverges from brain-like computation. To address these issues, we propose the Spiking STDP Transformer (S$^{2}$TDPT), a neuromorphic Transformer that implements self-attention through spike-timing-dependent plasticity (STDP), embedding query--key correlations in synaptic weights. STDP, a core mechanism of memory and learning in the brain and widely studied in neuromorphic devices, naturally enables in-memory computing and supports non-von Neumann hardware. On CIFAR-10 and CIFAR-100, our model achieves 94.35\% and 78.08\% accuracy with only four timesteps and 0.49 mJ on CIFAR-100, an 88.47\% energy reduction compared to a standard ANN Transformer. Grad-CAM shows that the model attends to semantically relevant regions, enhancing interpretability. Overall, S$^{2}$TDPT illustrates how biologically inspired attention can yield energy-efficient, hardware-friendly, and explainable neuromorphic models.

Attention via Synaptic Plasticity is All You Need: A Biologically Inspired Spiking Neuromorphic Transformer

arXiv:2511.08955v2 Announce Type: replace-cross 
Abstract: Simulating microstructure evolution (MicroEvo) is vital for materials design but demands high numerical accuracy, efficiency, and physical fidelity. Although recent studies on deep learning (DL) offer a promising alternative to traditional solvers, the field lacks standardized benchmarks. Existing studies are flawed due to a lack of comparing specialized MicroEvo DL models with state-of-the-art spatio-temporal architectures, an overemphasis on numerical accuracy over physical fidelity, and a failure to analyze error propagation over time. To address these gaps, we introduce MicroEvoEval, the first comprehensive benchmark for image-based microstructure evolution prediction. We evaluate 14 models, encompassing both domain-specific and general-purpose architectures, across four representative MicroEvo tasks with datasets specifically structured for both short- and long-term assessment. Our multi-faceted evaluation framework goes beyond numerical accuracy and computational cost, incorporating a curated set of structure-preserving metrics to assess physical fidelity. Our extensive evaluations yield several key insights. Notably, we find that modern architectures (e.g., VMamba), not only achieve superior long-term stability and physical fidelity but also operate with an order-of-magnitude greater computational efficiency. The results highlight the necessity of holistic evaluation and identify these modern architectures as a highly promising direction for developing efficient and reliable surrogate models in data-driven materials science.

MicroEvoEval: A Systematic Evaluation Framework for Image-Based Microstructure Evolution Prediction

arXiv:2511.08986v2 Announce Type: replace 
Abstract: Randomized controlled trials (RCTs) are indispensable for establishing the clinical value of medical artificial-intelligence (AI) tools, yet their high cost and long timelines hinder timely validation as new models emerge rapidly. Here, we propose BRIDGE, a data-reuse RCT design for AI-based risk models. AI risk models support a broad range of interventions, including screening, treatment selection, and clinical alerts. BRIDGE trials recycle participant-level data from completed trials of AI models when legacy and updated models make concordant predictions, thereby reducing the enrollment requirement for subsequent trials. We provide a practical checklist for investigators to assess whether reusing data from previous trials allows for valid causal inference and preserves type I error. Using real-world datasets across breast cancer, cardiovascular disease, and sepsis, we demonstrate concordance between successive AI models, with up to 64.8% overlap in top 5% high-risk cohorts. We then simulate a series of breast cancer screening studies, where our design reduced required enrollment by 46.6%--saving over US$2.8 million--while maintaining 80% power. By transforming trials into adaptive, modular studies, our proposed design makes Level I evidence generation feasible for every model iteration, thereby accelerating cost-effective translation of AI into routine care.

تعتبر التجارب السريرية العشوائية (RCTs) ضرورية لتأكيد الفعالية السريرية لأدوات الذكاء الاصطناعي الطبية، لكن تكاليفها العالية والجداول الزمنية الطويلة تشكل تحديات كبيرة. يقدم التصميم المقترح BRIDGE حلاً من خلال إعادة استخدام بيانات المشاركين من التجارب السابقة عندما تنتج نماذج الذكاء الاصطناعي تنبؤات متشابهة. يمكن أن يقلل هذا النهج من متطلبات التسجيل بنسبة 46.6% ويوفر أكثر من 2.8 مليون دولار مع الحفاظ على قوة إحصائية تبلغ 80%، مما يبرز إمكانيته في التحقق الفعال من نماذج الذكاء الاصطناعي في مجالات مثل سرطان الثدي وأمراض القلب والإنتان.

Los ensayos controlados aleatorios (ECA) son esenciales para validar la efectividad clínica de las herramientas de IA médica, pero sus altos costos y largos plazos presentan desafíos significativos. El diseño propuesto BRIDGE ofrece una solución al reutilizar datos de participantes de ensayos anteriores cuando los modelos de IA producen predicciones similares. Este enfoque puede reducir los requisitos de inscripción en un 46.6% y ahorrar más de 2.8 millones de dólares, manteniendo una potencia estadística del 80%, demostrando su potencial para una validación eficiente de modelos de IA en áreas…

Les essais contrôlés randomisés (ECR) sont essentiels pour valider l'efficacité clinique des outils d'IA médicale, mais leurs coûts élevés et leurs délais prolongés posent des défis importants. La conception proposée BRIDGE offre une solution en réutilisant les données des participants d'essais précédents lorsque les modèles d'IA produisent des prédictions similaires. Cette approche peut réduire les besoins d'inscription de 46,6 % et économiser plus de 2,8 millions de dollars tout en maintenant une puissance statistique de 80 %, démontrant son potentiel pour une validation efficace des modèles…

Randomized controlled trials (RCTs) are essential for validating the clinical effectiveness of medical AI tools, but their high costs and lengthy timelines pose significant challenges. The proposed BRIDGE design offers a solution by reusing participant-level data from previous trials when AI models yield similar predictions. This approach can significantly reduce enrollment requirements by 46.6% and save over $2.8 million while maintaining an 80% statistical power, demonstrating its potential for efficient AI model validation in areas like breast cancer, cardiovascular disease, and sepsis.

Distribution-Based Feature Attribution for Explaining the Predictions of Any Classifier

Was this article worth reading? Share it