Moral robustness R is mostly determined in post-training because it varies systematically by model family

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Moral Robustness (R)concept0.832
  Metric measuring within-persona stability of MFQ responses; formalizes model consistency when simulating a given character
• Moral susceptibility S is largely shaped by pre-training because it shows low cross-model variance not predicted by model familyhypothesis0.821
  Theoretical interpretation of the empirical cross-model variance pattern for S
• Secure fine-tuning largely preserves the base moral foundations profile, showing profile saturation is specific to misalignment-inducing trainingfinding0.779
  Control comparison confirming ceiling shift is not a generic fine-tuning artifact
• Does robust agency suffice for moral patienthood?question0.774
  Key normative question for the agency route.
• Robust agency suffices for moral patienthood.claim0.766
  Normative premise of the robust agency route.
• Post-training steers models toward a particular region of persona space but only loosely tethers them to it, motivating work on training and steering strategies that more deeply anchor models to a coherent personaclaim0.759
  Central interpretive claim and motivation for future work
• Is there a tradeoff between subtlety of trait expression and robustness in character-trained models?question0.759
  Open question raised in Appendix E regarding misaligned persona behavior
• How does different post-training data shift a model's position along persona dimensions?question0.753
  Future work direction: using persona space to study effects of training data on model character