{"ID":2877641,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.00091","arxiv_id":"2509.00091","title":"Ensemble Debates with Local Large Language Models for AI Alignment","abstract":"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.","short_abstract":"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...","url_abs":"https://arxiv.org/abs/2509.00091","url_pdf":"https://arxiv.org/pdf/2509.00091v2","authors":"[\"Ephraiem Sarabamoun\"]","published":"2025-08-27T13:25:51Z","proceeding":"cs.AI","tasks":"[\"cs.AI\",\"cs.CL\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}