INDEX

Explanations

data formats and structured data

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

werben

1.46

égi

1.44

ürlich

1.37

 Grüße

1.34

agaman

1.32

Württemberg

1.24

 šal

1.24

]].

1.21

</tfoot>

1.21

 blacklist

1.20

POSITIVE LOGITS

 whereas

1.40

 although

1.36

but

1.27

 exposure

1.24

 however

1.23

 unless

1.22

 though

1.19

as

1.18

 exposures

1.17

 while

1.15

Activations Density 0.251%