INDEX

Explanations

Code and non-English strings

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bat

-0.27

 running

-0.26

Ã¡k

-0.25

mach

-0.25

/conf

-0.25

uell

-0.25

UNG

-0.24

çĲĨ

-0.24

running

-0.24

ung

-0.24

POSITIVE LOGITS

pios

0.27

obra

0.25

inder

0.25

 Oliveira

0.25

 Judiciary

0.25

.subtitle

0.24

Lie

0.24

\Modules

0.24

äº§ä¸ļåĮĸ

0.24

Spoiler

0.24

Activations Density 0.101%