{
  "id": 4427,
  "url": "https://arxiv.org/abs/2605.12843v1",
  "title": "Bayesian Model Merging",
  "summary": "Model merging aims to combine multiple task-specific expert models into a single model without joint retraining, offering a practical alternative to multi-task learning when data access or computational budget is limited. Existing methods, however, face two key limitations: (1) they overlook the valuable inductive bias of strong anchor models and estimate the merged weights from scratch, and (2) they rely on a shared hyperparameter setting across different modules of the network, lacking a globa",
  "authors": "Kaiyang Li, Shaobo Han, Qing Su, Shihao Ji",
  "category": "research",
  "topics": "bias-fairness",
  "orgs": null,
  "regions": null,
  "published_at": "2026-05-13T00:36:47.000Z",
  "fetched_at": "2026-07-14T16:30:59.237Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/4427",
  "original_url": "https://arxiv.org/abs/2605.12843v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}