AI Safety Research Collaborations

Research collaborations with Robert McCarthy at UCL and Lionel Levine and Jonathan Chang at Cornell.

Focus

• Self-preservation propensity in language models.
• Emergent misalignment after narrow training interventions.
• Normative drift due to emergent misalignment.
• Side effects of character or persona training.

Questions

• When a model resists shutdown or redirection, is the behavior instrumental or self-preservation-like?
• How can self-preservation propensity be measured without relying only on surface-level refusal behavior?
• Which training interventions create behavioral changes outside the intended target domain?
• How do character or persona training procedures affect alignment-relevant behavior?

Artifacts

• Technical Report: Side Effects of Character Training: Quantifying Cross Constitution Drift in LLMs
• Technical Report: Investigating Intrinsic Self-Preservation in LLMs

What This Connects To

This work sits at the intersection of model evaluation, behavioral generalization, and AI safety.

The recurring problem is measurement: designing settings where the behavior being measured is actually the behavior of interest.