INDEX

Explanations

discussing or talking

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

if

-1.60

by

-1.47

to

-1.37

 추가

-1.23

 경우

-1.23

媼

-1.19

 XNUMX

-1.19

ոյ

-1.16

 حتی

-1.14

 embaj

-1.13

POSITIVE LOGITS

Finally

1.36



1.30

Akhir

1.29

 unut

1.28

 étrange

1.21

Sumber

1.16

 perles

1.14

Pada

1.13

 appels

1.13

歳

1.11

Activations Density 0.086%