Fine-tuning as character formation: what kinds of selves are produced through training is an open research direction.

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

• Fine-tuning is universal in increasing the strength of helpful and harmless intentions across model families.claim0.811
  Conclusion from Experiment 3 and HH intent analysis.
• What is the relationship between character and capability in fine-tuned models?question0.807
  Open question raised in Discussion noting capability changes could be due to character training itself, LoRA, or other factors
• Fine-Tuning via Reinforcement Learningmethod0.807
  Technique used to impose guardrails on base LLMs, analogized to censorship on the simulator's range of simulacra
• Character training for misalignment produces more subtle and realistic malicious behavior than fine-tuning on insecure code or activation steeringclaim0.805
  Realism advantage claim with safety implications; supported by anecdotal comparison in Appendix D
• SOO fine-tuning could be extended to align AI representations of its own goals with human user preferences, reducing misalignment by fostering coherence between self-related and other-related preferenceshypothesis0.802
  Future work hypothesis about extending SOO to direct value alignment
• Fine-tuning induces the behavioral pattern of self-correction but does not improve the underlying ability to correct effectivelyclaim0.799
  Key interpretive conclusion from the dissociation between attempt rate and improvement rate in fine-tuning experiments
• S suggests practical diagnostics for prompt design, retrieval, and light fine-tuning without additional training infrastructure.claim0.798
  Applied contribution.
• Fine Tuning and Adaptationconcept0.797
  The patient, hand-guided adjustment of shape and dimension to each unique condition in a building; requires materials that make it economical and easy.