INDEX

Explanations

comparison of two things

This neuron responds to comparative constructions—especially the pattern “than it did” (and its surrounding comparative context).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

dır

0.78

 عندك

0.74

ก็จะ

0.74

ـ

0.72

CDF

0.72

 surroundings

0.71

 coch

0.70

–

0.70

tone

0.70

 maybe

0.70

POSITIVE LOGITS

Than

1.03

 Than

0.95

than

0.90

 than

0.86

atual

0.79

 decât

0.77

头

0.71

張り

0.70

妃

0.66

entes

0.64

Activations Density 0.049%