{
  "id": 5413,
  "url": "https://arxiv.org/abs/2604.22891v4",
  "title": "Quantifying and Mitigating Self-Preference Bias of LLM Judges",
  "summary": "LLM-as-a-Judge has become a dominant approach in automated evaluation systems, playing critical roles in model alignment, leaderboard construction, quality control, and so on. However, the scalability and trustworthiness of this approach can be substantially distorted by Self-Preference Bias (SPB), which is a directional evaluative deviation in which LLMs systematically favor or disfavor their own generated outputs during evaluation. Existing measurements rely on costly human annotations and con",
  "authors": "Jinming Yang, Zheng Hu, Chuxian Qiu, Zhenyu Deng, Xinshan Jiao, Tao Zhou",
  "category": "research",
  "topics": "bias-fairness,safety-alignment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-04-24T09:46:22.000Z",
  "fetched_at": "2026-07-14T16:31:44.624Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/5413",
  "original_url": "https://arxiv.org/abs/2604.22891v4",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}