INDEX

Explanations

turning to face someone or something

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 VECTOR

1.35

 Ángel

1.29

 Canarias

1.26

 isotropic

1.25

 Puro

1.23

Muz

1.19

irit

1.18

Miguel

1.17

irtual

1.17

 Fres

1.16

POSITIVE LOGITS

Dinner

1.34

 dinner

1.34

 Dinner

1.32

lery

1.29

 Supper

1.28

 dinn

1.28

 Nancy

1.27

dinner

1.26

 Lacey

1.25

Rye

1.24

Activations Density 0.134%

No Known Activations