INDEX

Explanations

file extensions and code directives

The neuron strongly activates on tokens that are floating‐point number literals (i.e. numeric tokens containing decimal points).

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

يا

0.82

 ими

0.79

Sas

0.79

 mila

0.77

also

0.76

 Igre

0.76

 εἰ

0.76

mor

0.76

0.75

 eram

0.74

POSITIVE LOGITS

ార

0.85

文本

0.82

 swollen

0.75

鼠标

0.73

集合

0.72

リット

0.72

ACCESS

0.71

itôt

0.71

 hitter

0.71

乘以

0.71

Activations Density 0.000%