INDEX

Explanations

making informed decisions

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

to

-1.55

 که

-1.34

-1.31

{'

-1.22

 что

-1.16

𝟒

-1.16

%;

-1.15

metology

-1.14

isMethod

-1.14

いかが

-1.14

POSITIVE LOGITS

1.31

 hesit

1.30

on

1.27

 competir

1.22

шать

1.20

 بشأن

1.20

 impulsar

1.17

and

1.16

</em>

1.16

 hinzufügen

1.15

Activations Density 0.067%