INDEX

Explanations

conditional execution

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

$+

1.23

slightly

1.14

₁+

1.10

 marginally

1.10

 slightly

1.08

 barely

1.06

 légèrement

1.05

$+$

1.03

+{

1.01

 ligeramente

1.00

POSITIVE LOGITS

 thì

1.52

 allora

1.45

 then

1.32

 maka

1.31

 entonces

1.29

だったら

1.28

이라면

1.26

 τότε

1.25

ならば

1.23

 তাহলে

1.23

Activations Density 0.034%