INDEX

Explanations

attenuation or cause precedes effect

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

岖

2.45

</img>

2.28

 Compose

2.23

ATCH

2.20

ty

2.19

litz

2.17

 equid

2.13

 розташо

2.10

panic

2.05

चीत

2.01

POSITIVE LOGITS

yı

2.55

тво

2.50

 primaria

2.48

lovl

2.45

iš

2.40

 sürü

2.36

ہار

2.35

्ज

2.34

 साथी

2.33

ení

2.31

Activations Density 0.064%