2 papers
cs.LG2026
Constitutional adapters: Inference-time interventions for misalignment and misuse
Adam S. Lowet, Mark Kurzeja
Training models to act in accordance with an explicitly defined set of principles, or "constitution," has shown promise as a robust and transparent mechanism for AI alignment. Howe…
cs.LG2025
Consistency Training Helps Stop Sycophancy and Jailbreaks
Alex Irpan, Alexander Matt Turner, Mark Kurzeja +2
An LLM's factuality and refusal training can be compromised by simple changes to a prompt. Models often adopt user beliefs (sycophancy) or satisfy inappropriate requests which are…