arXiv subjects

Ephraiem Sarabamoun

Publications and source records attributed to Ephraiem Sarabamoun.

3 recordsLinked to original sources

arXiv2025

Special-Character Adversarial Attacks on Open-Source Language Model

Large language models (LLMs) have achieved remarkable performance across diverse natural language processing tasks, yet their vulnerability to character-level adversarial manipulations presents significant security challenges for real-world deployments. This paper presents a study of different special character attacks including unicode, homoglyph, structural, and textual encoding attacks aimed at bypassing safety mechanisms. We evaluate seven prominent open-source models ranging from 3.8B to 32B parameters on 4,000+ attack attempts. These experiments reveal critical vulnerabilities across all model sizes, exposing failure modes that include successful jailbreaks, incoherent outputs, and unrelated hallucinations.

cs.CR↗

arXiv2025

Ensemble Debates with Local Large Language Models for AI Alignment

As large language models (LLMs) take on greater roles in high-stakes decisions, alignment with human values is essential. Reliance on proprietary APIs limits reproducibility and broad participation. We study whether local open-source ensemble debates can improve alignmentoriented reasoning. Across 150 debates spanning 15 scenarios and five ensemble configurations, ensembles outperform single-model baselines on a 7-point rubric (overall: 3.48 vs. 3.13), with the largest gains in reasoning depth (+19.4%) and argument quality (+34.1%). Improvements are strongest for truthfulness (+1.25 points) and human enhancement (+0.80). We provide code, prompts, and a debate data set, providing an accessible and reproducible foundation for ensemble-based alignment evaluation.

cs.AI↗

arXiv2020

Solitons in the Einstein universe

We show that equations of Newtonian hydrodynamics and gravity with Einstein's cosmological constant included admit gravitostatic wave solutions propagating in the background of Einstein's static Universe. In the zero pressure limit these waves exist at an average matter density exceeding that of Einstein's Universe. They have the form of a lattice of integrable density singularities localized at the maxima of the gravitational potential. These singularities are steady-state counterparts of the so-called Zeldovich pancakes (ZP), interim wall-like structures appearing at nonlinear stages of development of gravitational instability. As the average matter density decreases, the period of the ZP lattice increases diverging at the density of Einstein's Universe. Solitary wave solutions are found at exactly the density of Einstein's Universe, and at a slightly larger density the wave may be viewed as a lattice of well-separated ZP solitons.

astro-ph.CO↗