INDEX

Explanations

purpose, contact, worst

The neuron fires on numeric tokens containing decimal points (floating-point numbers).

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

იმ

0.63

 यूज्ड

0.54

bolo

0.54

krb

0.53

amers

0.52

なかった

0.51

 Flush

0.50

हिले

0.49

心里

0.48

amt

0.48

POSITIVE LOGITS

 Intens

0.56

 intens

0.54

數

0.54

umpy

0.53

Wal

0.53

 staged

0.53

 proportionality

0.53

简化

0.53

顛

0.52

Simpl

0.51

Activations Density 0.000%