INDEX

Explanations

change in

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Latvia

1.88

 adjud

1.78

 Burkina

1.78

 Ethiopia

1.77

Akk

1.71

 Montenegro

1.70

ApiException

1.70

 Bulgaria

1.70

 Buenos

1.69

 Dör

1.69

POSITIVE LOGITS

幅度

1.78

over

1.69

 произошло

1.59

 signifikan

1.43

下來

1.41

傾向

1.38

home

1.36

feels

1.35

 alami

1.34

overs

1.33

Activations Density 0.082%

No Known Activations