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

arXiv:2602.15927v1 Announce Type: new Abstract: Generative large vision-language models (LVLMs) have recently achieved impressive performance gains, and their user base is growing rapidly. However, the security of LVLMs, in particular in a long-context multi-turn setting, is largely underexplored. In this paper, we consider the realistic scenario in which an attacker uploads a manipulated image to the web/social media. A benign user downloads this image and uses it as input to the LVLM. Our novel stealthy Visual Memory Injection […]

Ver mais

arXiv:2601.10825v1 Announce Type: new Abstract: Large language models have achieved remarkable capabilities across domains, yet mechanisms underlying sophisticated reasoning remain elusive. Recent reasoning models outperform comparable instruction-tuned models on complex cognitive tasks, attributed to extended computation through longer chains of thought. Here we show that enhanced reasoning emerges not from extended computation alone, but from simulating multi-agent-like interactions — a society of thought — which enables diversification and debate among internal cognitive perspectives characterized by distinct personality traits […]

Ver mais

arXiv:2602.11241v1 Announce Type: new Abstract: Self-play has enabled large language models to autonomously improve through self-generated challenges. However, existing self-play methods for vision-language models rely on passive interaction with static image collections, resulting in strong dependence on initial datasets and inefficient learning. Without the ability to actively seek visual data tailored to their evolving capabilities, agents waste computational effort on samples that are either trivial or beyond their current skill level. To address these limitations, we propose Active-Zero, […]

Ver mais

We introduce Strategic Doctrine Language Models (sdLM), a learning-system framework for multi-document strategic reasoning with doctrinal consistency constraints and calibrated uncertainty. The approach combines multi-document attention, temporal encoding, and a doctrine-consistency layer to improve long-horizon forecasting and plan plausibility while reducing severe doctrinal violations. We evaluate sdLM using (i) expert-panel scoring of strategic scenarios (N=47), (ii) doctrine consistency on 336 doctrine publications (12,847 statements), and (iii) geopolitical forecasting on 127 historical counterfactuals (1945-2020) across 12-60 month horizons. Across […]

Ver mais

arXiv:2510.06028v2 Announce Type: replace-cross Abstract: This paper provides data-dependent bounds on the expected error of the Gibbs algorithm in the overparameterized interpolation regime, where low training errors are also obtained for impossible data, such as random labels in classification. The results show that generalization in the low-temperature regime is already signaled by small training errors in the noisier high-temperature regime. The bounds are stable under approximation with Langevin Monte Carlo algorithms. The analysis motivates the design of an […]

Ver mais

In side-channel analysis, simple power analysis (SPA) is a widely used technique for recovering secret information by exploiting differences between operations in traces. However, in realistic measurement environments, SPA is often hindered by noise, temporal misalignment, and weak or transient leakage, which obscure secret-dependent features in single or very few power traces. In this paper, we provide a systematic analysis of moving-skewness-based trace preprocessing for enhancing asymmetric leakage characteristics relevant to SPA. The method computes local skewness within […]

Ver mais

Hey everybody, I work for a firm on an engineering team that uses AWS. Historically they’ve used PySpark to deploy deep loading models that I’ve built, but I’ve been tasked with researching to see if there’s a way to call models for inference as they say there is a decent amount of overhead as they are transitioning to a new mode of operation. They are running a spark cluster with around 300 nodes, and ultimately hope there is […]

Ver mais

Board games have long served as complex decision-making benchmarks in artificial intelligence. In this field, search-based reinforcement learning methods such as AlphaZero have achieved remarkable success. However, their significant computational demands have been pointed out as barriers to their reproducibility. In this study, we propose a model-free reinforcement learning algorithm designed for board games to achieve more efficient learning. To validate the efficiency of the proposed method, we conducted comprehensive experiments on five board games: Animal Shogi, Gardner […]

Ver mais