Trait-Subspace Neutralization reduces trait-induced instability in LLM refusal behavior by anchoring trait-conditioned representations to the no-trait baseline within a low-dimensional subspace.
Let’s verify step by step
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Making Your LLMs More Objective: Stabilizing LLM Safety Behavior Across Traits with Trait-Invariant Safety Tuning
Trait-Subspace Neutralization reduces trait-induced instability in LLM refusal behavior by anchoring trait-conditioned representations to the no-trait baseline within a low-dimensional subspace.