arXiv:2603.05618v1 Announce Type: new Abstract: Chain-of-Thought (CoT) prompting improves LLM reasoning but can increase privacy risk by resurfacing personally identifiable information (PII) from the prompt into reasoning traces and outputs, even under policies that instruct the model not to restate PII. We study such direct, inference-time PII leakage using a model-agnostic framework that (i) defines leakage as risk-weighted, token-level events across 11 PII types, (ii) traces leakage curves as a function of the allowed CoT budget, and (iii) […]

Ver mais

Skyjo is a simple stochastic card game with partial information, local replacement decisions, and score-reducing column removal events. This paper develops a formal mathematical model of the game, derives expected-score rules for turn-level actions, proves several dominance and threshold results, and evaluates a family of heuristic strategies through Monte Carlo simulation. The focus here lies on local optimality under explicit belief assumptions rather than a full equilibrium solution of the multiplayer game. Finally a simulation code is provided […]

Ver mais

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 and how their relationships evolve as training progresses. In this work, we present GRAPHIC, an architecture-agnostic approach that analyzes neural networks on a class level. It leverages confusion matrices derived from intermediate layers using linear classifiers. We interpret these […]

Ver mais

Following the pivotal success of learning strategies to win at tasks, solely by interacting with an environment without any supervision, agents have gained the ability to make sequential decisions in complex MDPs. Yet, reinforcement learning policies face exponentially growing state spaces in high dimensional MDPs resulting in a dichotomy between computational complexity and policy success. In our paper we focus on the agent’s interaction with the environment in a high-dimensional MDP during the learning phase and we introduce […]

Ver mais

arXiv:2603.00538v1 Announce Type: new Abstract: This paper introduces a new method for performing field transfer operations in black-box coupling, when source discretization information is not available. This approach uses a stochastic approximation of the Galerkin projection which leads to a method that asymptotically provides conservation. Error in the accuracy and conservation has been compared to the mesh intersection method and radial basis functions on a simple domain, as well as on meshes of the LTX fusion reactor. For […]

Ver mais

There’s a question that quietly predicts whether an AI agent will feel sharp or strangely distant: How long does the agent remain misaligned when a new domain term enters the conversation? It sounds almost trivial — a glossary issue, a maintenance task. But in practice, this single latency window reveals more about an agent’s reliability than most teams realize. Because language doesn’t evolve politely. It arrives mid‑conversation. A finance analyst introduces a new internal metric. A product team […]

Ver mais

arXiv:2601.17265v1 Announce Type: cross Abstract: The grade of membership model is a flexible latent variable model for analyzing multivariate categorical data through individual-level mixed membership scores. In many modern applications, auxiliary covariates are collected alongside responses and encode information about the same latent structure. Traditional approaches to incorporating such covariates typically rely on fully specified joint likelihoods, which are computationally intensive and sensitive to misspecification. We introduce a covariate-assisted grade of membership model that integrates response and covariate […]

Ver mais

arXiv:2604.01440v1 Announce Type: new Abstract: The shift toward IoT-enabled, sensor-driven systems has transformed how operational data is generated, favoring continuous, real-time event streams (ES) over static event logs. This evolution presents new challenges for Streaming Process Mining (SPM), which must cope with out-of-order events, concurrent activities, incomplete cases, and concept drifts. Yet, the evaluation of SPM algorithms remains rooted in outdated practices, relying on static logs or artificially streamified data that fail to reflect the complexities of real-world […]

Ver mais