EMBER augments existing erasure methods by precisely removing concept features from embeddings via sparse matrix factorization, cutting relearning recovery to 35% on Llama-3.1-8B from 70-76%.
and Rubinstein, Benjamin I
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Don't Forget Your Embeddings: Robust Knowledge Erasure via Precise Editing of Embeddings
EMBER augments existing erasure methods by precisely removing concept features from embeddings via sparse matrix factorization, cutting relearning recovery to 35% on Llama-3.1-8B from 70-76%.