INDEX

Explanations

variability and streams

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 poffe

-0.91

 myſelf

-0.90

Efq

-0.89

 itſelf

-0.88

 reaſon

-0.81

 poffible

-0.79

 themſelves

-0.78

 Majefty

-0.78

 ſtand

-0.78

 pleaſure

-0.77

POSITIVE LOGITS

0.90

ers

0.74

ron

0.69

ry

0.66

ual

0.62

sun

0.61

0.60

ings

0.60

ative

0.59

son

0.58

Activations Density 0.189%