INDEX

Explanations

how to implement in code

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

 якщо

-1.73

لباس

-1.45

如果

-1.39

 лікування

-1.36

 producing

-1.35

if

-1.34

If

-1.30

It

-1.29

 controlling

-1.24

太多了

-1.24

POSITIVE LOGITS

廊下

1.46

melons

1.44

 actionTypes

1.33

 this

1.27

innen

1.27

はこの

1.25

jetos

1.24

blume

1.24

 neder

1.22

 einerseits

1.22

Activations Density 0.015%