conftrace_

invariant risk minimization (irm)

1 paper

Explore in graph

Co-occurring keywords · papers mentioning both

hazardous knowledge (1) unlearning (1) downstream fine-tuning (1) large language model (1) invariance (1)

Papers

Invariance Makes LLM Unlearning Resilient Even to Unanticipated Downstream Fine-Tuning ICML 2025