@misc{indiciaef8583c7d5e21,
  title = {Compositional Concept Erasure in Text-to-Image Diffusion Models via Hierarchically Grounded Semantic Surgery},
  author = {Chen Dai and Ganyu Zou and Nathan Self and Kevin Piper and Ramachandra Rao Seethiraju and Karthik Shyamsunder and Chang-Tien Lu and Naren Ramakrishnan},
  year = {2026},
  url = {https://arxiv.org/abs/2610.07337},
  note = {Source identifier: 2610.07337}
}