INDEX

Explanations

questions, commands, or code definitions

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

 only

-1.86

in

-1.76

to

-1.67

on

-1.50

of

-1.44

so

-1.41

 just

-1.38

 ONLY

-1.34

for

-1.30

POSITIVE LOGITS

鹛

1.33

 probado

1.31

⓹

1.27

 herido

1.23

 TOWNS

1.21

bleshooting

1.21

栊

1.21

 empêcher

1.18

㈩

1.17

layah

1.16

Activations Density 0.001%