INDEX

Explanations

in followed by context

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

was

-1.51

くて

-1.32

 always

-1.27

tanleria

-1.27

 could

-1.23

ऑ

-1.21

 thoſe

-1.17

ро

-1.17

这是一

-1.16

 malheure

-1.14

POSITIVE LOGITS

჻

1.69

 zahlreichen

1.55

なの

1.53

 molti

1.49

だったが

1.43

隂

1.42

 différents

1.42

 vielfäl

1.41

 diversos

1.39

 reiterated

1.38

Activations Density 0.039%