conftrace_

orthogonal direction

2 papers

Explore in graph

Co-occurring keywords · papers mentioning both

structured optimization (1) multidimensional analysis (1) trigger token (1) finite difference method (1) representation shift (1) refusal behavior (1) zeroth-order optimization (1) nonsmooth optimization (1) finite difference (1) safety alignment (1)

Papers

The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions ICML 2025 An Optimal Structured Zeroth-order Algorithm for Non-smooth Optimization NeurIPS 2023