KL Divergence

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

• KL Divergence Retention Evaluationmethod0.869
  Measuring KL divergence between original and post-intervention outputs on Alpaca prompts to assess behavioral preservation
• Representational Divergenceconcept0.809
  Core phenomenon studied: when causal interventions shift internal representations away from the natural distribution
• VS-Standard achieves KL divergence of 0.12-0.13 from pretraining distribution on US state naming task, vs. 14.89-16.16 for Direct promptingfinding0.767
  Strong empirical evidence that VS recovers pretraining distribution while direct prompting collapses
• VS-Standard achieves KL divergence of 0.54 from pretraining distribution on Open-Ended QA, vs. 3.14 for Direct and 0.58 for Sequencefinding0.758
  Shows VS substantially better approximates the pretraining distribution than baseline methods
• KL Controlframework0.757
  Control framework minimizing expected complexity; shown to be a special case of expected free energy minimization
• Harmless Divergenceconcept0.751
  Divergences that occur in the behavioral null-space and do not affect functional claims about the model
• Pernicious Divergenceconcept0.747
  Divergences that activate hidden pathways or cause dormant behavioral changes, undermining mechanistic claims
• VS-Standard achieves KL divergence of 0.027 from uniform distribution on dice roll simulation, vs. 0.926 for Direct promptingfinding0.718
  Shows VS enables LLMs to better approximate random behavior compared to direct prompting