INDEX

Explanations

<start_of_turn> model

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

$\\

1.59

 nbsp

1.37

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.32

ichtung

1.31

↵↵↵↵↵↵↵

1.27

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.26

 sàn

1.25

 lực

1.24

↵↵↵↵↵

1.23

↵↵↵↵↵↵

1.22

POSITIVE LOGITS

</h2>

1.42

 otherwise

1.33

 wade

1.32

 matchmaking

1.31

 cuatro

1.29

 towards

1.27

 newest

1.27

 starred

1.20

 toward

1.20

Brand

1.19

Activations Density 0.435%