{
  "id": 3858,
  "url": "https://arxiv.org/abs/2605.23780v1",
  "title": "Beyond Binary Edits Robust Multimodal Knowledge Editing with Adversarial Subspace Alignment",
  "summary": "Multimodal large language models (MLLMs) need efficient mechanisms to update knowledge without degrading existing capabilities. While intrinsic multimodal knowledge editing achieves strong reliability and locality, it often exhibits limited generality, failing to propagate edits across semantically equivalent visual and linguistic variations. This issue arises from the lack of explicit semantic supervision, rigid editing scopes, and biased anchoring to individual samples in high-dimensional mult",
  "authors": "Haoyuan Wang, Xiaohao Liu, Jiajie Su, Jianmao Xiao, Chaochao Chen",
  "category": "research",
  "topics": "bias-fairness,safety-alignment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-05-22T15:46:10.000Z",
  "fetched_at": "2026-07-14T16:30:31.923Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/3858",
  "original_url": "https://arxiv.org/abs/2605.23780v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}