Norm-Anchor Scaling breaks the norm-feedback loop in sequential LLM editing by anchoring value vectors to original norms, improving long-run performance by 72.2% and extending the editing horizon over 4x.
tilde” versions: E h ∥vnew n ∥2 ∥ ˜Wn−1∥2 i ≈s new∥ ˜Wn−1∥2+bnew,E h ∥vold n ∥2 ∥ ˜Wn−1∥2 i ≈s old∥ ˜Wn−1∥2+bold. Under these empirical observations for the “tilde
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Norm Anchors Make Model Edits Last
Norm-Anchor Scaling breaks the norm-feedback loop in sequential LLM editing by anchoring value vectors to original norms, improving long-run performance by 72.2% and extending the editing horizon over 4x.