INDEX

Explanations

feeling afraid or scared

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

かと思います

-0.98

 falsche

-0.89

 Kien

-0.87

擾

-0.84

壶

-0.83

järvi

-0.83

……….

-0.82

 ごはん

-0.81

档

-0.81

 organes

-0.77

POSITIVE LOGITS

of

3.13

to

2.02

 afraid

1.41

ที่จะ

1.24

 Afraid

1.23

or

1.08

 Fear

1.07

 that

1.05

 takut

0.98

 paura

0.93

Activations Density 0.010%