Base64 feature A/1/2357 and B/1/2165 have activation correlation of 0.85

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

• DNA feature A/1/2937 and B/1/3680 have activation correlation of 0.92finding0.865
  Universality of DNA feature across two transformer models with different random seeds
• Hebrew feature A/1/416 and B/1/1901 have activation correlation of 0.92finding0.859
  Universality of Hebrew script feature across two transformers
• Arabic feature A/1/3450 and B/1/1334 have activation correlation of 0.91 across 40M tokensfinding0.854
  Demonstrates universality of the Arabic script feature across two independently trained transformers
• Activating the base64 feature A/1/2357 causes the model to generate base64 textfinding0.821
  Causal validation of base64 feature function via pinned feature sampling
• Most correlated neuron A/neurons/470 has correlation of only 0.18 with base64 feature A/1/2357 and responds to code, HTML labels, URLsfinding0.812
  Shows base64 feature is polysemantic at neuron level but monosemantic as learned feature
• Feature pair A/1/3949 and B/1/3321 have activation correlation 0.98 but negative logit weight correlation, firing on PLOSOne journal abbreviationsfinding0.806
  Demonstrates that activation similarity can diverge from logit weight similarity due to interference
• Features in A/1 have median activation correlation of 0.72 with most similar feature in B/1; neurons have median 0.46finding0.805
  Systematic comparison showing features are substantially more universal than neurons across models
• Binarized DNA proxy has Pearson correlation of 0.80 with A/1/2937 feature activationsfinding0.787
  Demonstrates specificity and sensitivity of DNA feature