INDEX

Explanations

state of being

The neuron detects descriptive state or condition words—i.e. adjectives and adverbs that qualify or evaluate something’s status or readiness.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

if

-1.77

of

-1.72

by

-1.59

 навіть

-1.52

 when

-1.46

 Leur

-1.34

for

-1.32

to

-1.27

as

-1.25

Отзывы

-1.23

POSITIVE LOGITS

it

1.53

 favorable

1.31

 other

1.23

 ganhou

1.17

hs

1.15

ht

1.13

 conseguiu

1.11

 bén

1.10

vs

1.10

 abbiamo

1.09

Activations Density 0.052%