INDEX

Explanations

making things from parts

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

 they

-1.77

if

-1.41

but

-1.26

êtres

-1.23

 when

-1.16

wendige

-1.15

 können

-1.11

 podía

-1.10

},{

-1.10

 paar

-1.09

POSITIVE LOGITS

 from

1.57

 véhic

1.32

 obé

1.26

انا

1.26

 éta

1.23

 éprou

1.22

one

1.21

 Untersuch

1.20

 redé

1.19

 bestehende

1.16

Activations Density 0.212%