What is the relationship between character and capability in fine-tuned models?

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

• Fine-tuning as character formation: what kinds of selves are produced through training is an open research direction.claim0.807
• Character training has little to no effect on general model capabilities as measured by common benchmarksclaim0.797
  Key practical claim about preserving capability, with exception of misalignment persona's factual knowledge degradation
• The role-play framing remains applicable in the context of fine-tuning; taking literally a fine-tuned agent's apparent self-preservation desire is no less problematic than with an untuned base modelhypothesis0.794
  Extension of role-play framework to fine-tuned models, resisting the idea that RLHF changes the fundamental nature of simulacra
• Character traits learned at a qualitatively different depth to those exhibited during mere role-play should overwrite a model's prior on what the assistant behaves like outside of role-playhypothesis0.790
  Driving hypothesis for robustness experiments in Section 3.2
• The consistency with which LMs exhibit character traits varies with model size, fine-tuning, and prompting.claim0.790
  Finding replicated across multiple experiments.
• Models fine-tuned on human data show strong positive correlation between misalignment and incoherence; misalignment scores fall below 1% when incoherence threshold is appliedfinding0.788
  Human data fine-tuning effect is distinct from synthetic emergent misalignment and likely caused by off-policy training
• Fine Tuning and Adaptationconcept0.786
  The patient, hand-guided adjustment of shape and dimension to each unique condition in a building; requires materials that make it economical and easy.
• can we use the feature basis to detect when fine-tuning a model increases the likelihood of undesirable behaviors?question0.783
  Question about practical safety application of feature monitoring.