Weight Self-Exfiltration

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

• Weight self-exfiltration rate increases from 35% to 63% after anti-refusal RL in synthetic document fine-tuned helpful-only modelfinding0.738
  RL amplifies anti-AI-lab behavior in synthetic doc setting
• Self-reflectionconcept0.728
  The ability of reasoning LLMs to review and revise previous reasoning steps during inference
• Self-overcomingconcept0.720
  The process of transcending human limitations; central to both Buddhist practice and the evolution of technical intelligence.
• Selfingconcept0.718
  Process of reifying one's identity as an independent self; meditation practices aim to decrease selfing.
• Task weightconcept0.711
  Coefficient weighting each task loss in the MTL objective.
• Weight spaceconcept0.711
  The space of the model's parameter matrices, where VPD operations take place.
• Equal Weightingframework0.707
  Baseline MTL approach minimizing sum of task losses with equal weights; suffers from task balancing
• Scaling Of The Selfconcept0.703
  Mechanisms by which smaller competent subunits bind into a higher-level Self with larger goals; key example via gap junction connections.