{
  "id": 3453,
  "url": "https://arxiv.org/abs/2605.30873v1",
  "title": "Federated Variational Preference Alignment with Gumbel-Softmax Prior for Personalized User Preferences",
  "summary": "Federated Learning (FL) offers a privacy-preserving pathway for aligning Large Language Models (LLMs); however, existing frameworks typically enforce a monolithic reward model, inevitably averaging out inherently conflicting user preferences (e.g., helpfulness vs. harmlessness). While Variational Preference Learning (VPL) offers a pathway to personalization, adapting it to decentralized settings presents a fundamental challenge: posterior collapse driven by severe local data scarcity and heterog",
  "authors": "Jabin Koo, Hoyoung Kim, Minwoo Jang, Jungseul Ok",
  "category": "research",
  "topics": "safety-alignment,privacy-surveillance",
  "orgs": null,
  "regions": null,
  "published_at": "2026-05-29T05:52:21.000Z",
  "fetched_at": "2026-07-14T16:30:14.372Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/3453",
  "original_url": "https://arxiv.org/abs/2605.30873v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}