INDEX

Explanations

people, places, and specific entities

"blurred", "failure", "UN", "game", "thought"

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 suffix

0.49

 recharging

0.47

 rinsing

0.46

 facilitating

0.46

numer

0.46

ad

0.46

MOT

0.46

Ser

0.45

 suffixes

0.45

 affixed

0.44

POSITIVE LOGITS

yı

0.49

PROJECT

0.46

կ

0.45

"),(

0.44

嵋

0.44

 trama

0.44



0.43

ということで

0.43

CacheV

0.42

삶

0.42

Activations Density 0.001%