INDEX

Explanations

looking down on others

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

reen

-0.80

moda

-0.80

oka

-0.79

сен

-0.75

фом

-0.73

 Россий

-0.72

 Hwang

-0.72

 ngoài

-0.71

 スカート

-0.71

 privati

-0.71

POSITIVE LOGITS

 disdain

2.25

 contempt

2.09

 looked

1.93

 condescending

1.84

bel

1.80

瞧

1.77

 look

1.72

 down

1.70

 despise

1.69

 underestimate

1.67

Activations Density 0.027%