INDEX

Explanations

going to discuss experiences

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

 fluvial

1.32

 astronomical

1.27

 pilgrims

1.21

Corn

1.20

Prove

1.18

 crucifix

1.16

ocarbon

1.14

.;

1.13

Roman

1.12

 sandw

1.10

POSITIVE LOGITS

閥

2.19

tik

1.96

scht

1.84

 αυτές

1.79

 liberd

1.74

แนน

1.74

uridad

1.74

izlik

1.72

 pikiran

1.71

ləşdir

1.71

Activations Density 0.002%