INDEX

Explanations

speaker, Speakers

what the speaker says

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 जिस

0.38

видно

0.38

ಾದ

0.37

 exogenous

0.37

 gezeigt

0.37

 حمد

0.37

 utilizzato

0.37

 getter

0.36

မဟုတ်

0.36

芸

0.36

POSITIVE LOGITS

 speakers

2.16

 speaker

2.13

speaker

2.02

 Speakers

2.02

Speakers

1.98

Speaker

1.93

speakers

1.93

 Speaker

1.91

 SPEAKER

1.70

 स्पीकर

1.68

Activations Density 0.098%