INDEX

Explanations

comparison of or DAG followed by (

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

 אך

1.35

aing

1.33

רים

1.26

traf

1.24

 mentioning

1.23

 puisque

1.22

 খেয়ে

1.21

 condemning

1.20

 denying

1.18

oš

1.17

POSITIVE LOGITS

ங்கிணை

1.45

선

1.26

обходимо

1.23

$/.

1.19

可

1.18

 좋다

1.17

适合

1.16

نى

1.14

はお

1.10

ウド

1.10

Activations Density 0.026%