INDEX

Explanations

questions starting with if when

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

extremely

-1.29

ſſen

-1.27

 POSSIBLE

-1.25

 hemd

-1.23

representing

-1.22

weile

-1.21

grew

-1.21

hehehe

-1.20

 wana

-1.18

antaranya

-1.18

POSITIVE LOGITS

 your

2.58

you

1.68

1.38

use

1.32

1.31

an

1.23

as

1.20

 using

1.20

</b>

1.19

Activations Density 0.294%