Monotonic Scaling Property

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

• Introspective capacity may follow a simple monotonic scaling law across all concepts and architectureshypothesis0.760
  The paper treats this as possible but unconfirmed; current evidence shows concept-specific scaling only
• Multidimensional Scalingmethod0.752
  Used in the color cooccurrence experiment to embed colors into 3D space preserving dissimilarity matrix distances
• Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet (Templeton et al., 2024)concept0.747
  Key paper on scaling SAE-based interpretability to frontier models, cited as precedent
• Power law scalingconcept0.745
  Observation that SAE loss decreases as a power law with compute budget.
• Scaling Of The Selfconcept0.739
  Mechanisms by which smaller competent subunits bind into a higher-level Self with larger goals; key example via gap junction connections.
• Monotonicity Natural Language Inferenceconcept0.732
  NLI task where premise-hypothesis pairs differ by a single word replaced by hypernym/hyponym, with negation as a variable.
• entropy scalingconcept0.730
  How the entropy gain ΔS scales with perimeter length P
• Inverse Scaling Lawconcept0.726
  Hypothesis cited in paper suggesting deceptive capabilities may scale with model size