How Language Models Choose Sides: Internal Representations of Instruction Hierarchy
It is shown that user-preferring conflict resolution can coexist with a readable internal arbitration signal, and successful intervention depends on the geometry of the readout rather than probe accuracy alone, while directions selected mainly for pooled separability steer poorly.
Enrique Balp-Straffon, Chih-Hao Hsu, Rushiraj Gadhvi et al.
· 0 citations