arXiv preprints from January 1, 2026 through September 25, 2026 — 14:24:37 EST
Neural networks are growing more capable on their own, but we do not understand their neural mechanisms. Understanding these mechanisms' decision-making processes, or mechanistic interpretability, enables (1) accountability and control in high-stakes domains, (2) the study of digital brains and the emergence of cognition, and (3)...
Convolutional Neural Networks (CNNs) have demonstrated remarkable success in image classification tasks; however, the choice between designing a custom CNN from scratch and employing established pre-trained architectures remains an important practical consideration. In this work, we present a comparative analysis of a custom-designed...
Bayesian inference is a popular approach to calibrating uncertainties, but it can underpredict such uncertainties when model misspecification is present, impacting its reliability to inform decision making. Recently, the statistics and machine learning communities have developed prediction-oriented inference approaches that provide...
Dense retrieval models are typically fine-tuned with contrastive learning objectives that require binary relevance judgments, even though relevance is inherently graded. We analyze how graded relevance scores and the threshold used to convert them into binary labels affect multilingual dense retrieval. Using a multilingual dataset...
This paper develops a computational framework for Multi-Period Martingale Optimal Transport (MMOT), addressing convergence rates, algorithmic efficiency, and financial calibration. Our contributions include: (1) Theoretical analysis: We establish discrete convergence rates of $O(\sqrt{Δt} \log(1/Δt))$ via Donsker's principle and...
Human cognition, driven by complex neurochemical processes, oscillates between imagination and reality and learns to self-correct whenever such subtle drifts lead to hallucinations or unsafe associations. In recent years, LLMs have demonstrated remarkable performance in a wide range of tasks. However, they still lack human cognition...
Static word embeddings often absorb social biases from the text they learn from, and those biases can quietly shape downstream systems. Prior work that uses the Stereotype Content Model (SCM) has focused mostly on single-group bias along warmth and competence. We broaden that lens to intersectional bias by building compound...
This paper introduces a fuzzy reinforcement learning framework, Enhanced-FQL($λ$), that integrates novel Fuzzified Eligibility Traces (FET) and Segmented Experience Replay (SER) into fuzzy Q-learning with the Fuzzified Bellman Equation (FBE) for continuous control. The proposed approach employs an interpretable fuzzy rule base instead...
Image Phase Alignment Super-sampling (ImPASS) is a computational method for combining displaced low-resolution images into a single high-resolution image. The general steps include measuring the relative displacements, up-sampling, aligning and combining the images, followed by a blind deconvolution. Previous ImPASS studies have shown...
High-quality methodology figures are central to scientific communication, yet they remain difficult and time-consuming to create. Such figures must distill a method's components and information flow into a clear, revisable diagram as the paper evolves. Existing methodology diagram automation systems typically face a trade-off between...
Current safety evaluations of large language models (LLMs) create a dangerous illusion of universal protection by aggregating harms under generic categories such as "Identity Hate", obscuring vulnerabilities toward specific populations. In this work, we expose the Selective Safety Trap: a systemic failure mode where models robustly...
Language diffusion models aim to improve sampling speed and coherence over autoregressive LLMs. We introduce Neural Flow Diffusion Models for language generation, an extension of NFDM that enables the straightforward application of continuous diffusion models to discrete state spaces. NFDM learns a multivariate forward process from...
Clustering customer chat data is vital for cloud providers handling multi service queries. Traditional methods struggle with overlapping concerns and create broad, static clusters that degrade over time. Reclustering disrupts continuity, making issue tracking difficult. We propose an adaptive system that segments multi turn chats into...
Negotiation is a core component of social intelligence, requiring agents to balance strategic reasoning, cooperation, and social norms. Recent work shows that LLMs can engage in multi-turn negotiation, yet nearly all evaluations occur exclusively in English. Using controlled multi-agent simulations across Ultimatum, Buy-Sell, and...
The effects of magnetisation forces in a binary mixture of gases characterised by large differences in magnetic susceptibility are studied using numerical simulations, with a focus on the differential diffusion of the species and the role of the gradient of mixture composition on the flow field resulting from magnetically-induced...
We consider a fourth-order regularization of the curvature flow for an immersed plane curve with fixed boundary, using an elastica-type functional depending on a small positive parameter $\varepsilon$. We show that the approximating flow smoothly converges, as $\varepsilon \to 0^+$, to the curvature flow of the curve with Dirichlet...
Classical anti-concentration results focus on the random sum $S := \sum _{i=1}^n ξ_i v_i$, where $ξ_i$ are independent random variables and $v_i$ are real numbers. In this paper, we prove new concentration results concerning the random sum $S := \sum_{i=1}^n w_{π_i } v_i $, where $w_i , v_i$ are real numbers and $π$ is a random permutation.
In this article, we formulate and analyze a new non-linear mathematical model to describe the dynamics of unemployment with a discouraged working population. We consider five dynamic variables, namely, unskilled unemployed individuals, skilled unemployed individuals, discouraged individuals, employed persons, and job vacancies....
Many hypersurfaces in algebraic geometry, such as discriminants, arise as the projection of another variety. The real complement of such a hypersurface partitions its ambient space into open regions. In this paper, we propose a new method for computing these regions. Existing methods for computing regions require the explicit equation...
Many emerging many-core accelerators replace a single large device memory with hundreds to thousands of lightweight cores, each owning only a small local SRAM and exchanging data via explicit on-chip communication. This organization offers high aggregate bandwidth, but it breaks a key assumption behind many volumetric rendering...
Foundation models for vision are predominantly trained on RGB data, while many safety-critical applications rely on non-visible modalities such as infrared (IR) and synthetic aperture radar (SAR). We study whether a single flow-matching foundation model pre-trained primarily on RGB images can be repurposed as a cross-spectral...
Psychedelics have shown potential in treating a range of mental health conditions, yet far less is known about their impact on creativity. This study examined three components of creativity-divergent thinking, cognitive reflection, and insight in a large sample (N = 5,905) from the Great British Intelligence Test. We compared...
Gravitational waves from binary black hole mergers yield values for both the black hole remnant mass $M$ and it's spin $a$, with the $169$ $a$ values collected so far crowding significantly around their average $\bar{a}=0.6869\pm 0.087$. Could this crowding relate directly to the Davies phase transition point at $a=0.68125$ from black...
As artificial intelligence increasingly drives critical decisions, the ability to genuinely explain how neural networks make predictions is essential for trust. Yet, most current explanation methods offer post-hoc rationalizations rather than guaranteeing a true reflection of the model's reasoning. We introduce the notion of...
Retrieval-Augmented Generation (RAG) has emerged as an important means of enhancing the performance of large language models (LLMs) in knowledge-intensive tasks. However, most existing RAG strategies treat retrieved passages in a flat and unstructured way, which prevents the model from capturing structural cues and constrains its...