INDEX

Explanations

descriptive adjectives and nouns

This neuron activates on descriptive modifiers—especially adjectives (and some adverbs) that provide vivid qualities or states (e.g. fabulous, legendary, rackety, contaminated).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 famí

-1.41

 caricatura

-1.33

𖣔

-1.31

 花朵

-1.28

 ۱۸

-1.27

𓁹

-1.27

𒊹

-1.21

Destin

-1.20

Academia

-1.18

Hana

-1.16

POSITIVE LOGITS

If

1.13

 milhões

1.06

(!)

1.05

if

1.05

 Please

1.00

How

1.00

際に

0.97

ac

0.96

 hoher

0.96

Let

0.95

Activations Density 0.086%