INDEX

Explanations

only transfers differences

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Presentation

1.61

 الذي

1.59

 التي

1.54

 التى

1.50

 الاسئله

1.37

میں

1.35

 عباره

1.35

হইতে

1.34

dengan

1.33

على

1.30

POSITIVE LOGITS

 check

1.28

 bơm

1.27

 trumpet

1.24

 guard

1.23

 거고

1.23

 anyway

1.21

 trace

1.20

get

1.20

sum

1.19

 doubly

1.17

Activations Density 0.005%