INDEX

Explanations

cost and savings

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

稍

1.18

 régulièrement

1.17

不少

1.17

 extensively

1.16

 decently

1.06

 regularly

1.06

 преимущественно

1.05

陆续

1.03

 comprehensively

1.03

 लंबे

1.03

POSITIVE LOGITS

 مسئله

1.61

 dilemma

1.52

 مساله

1.51

 Frage

1.49

 questão

1.46

 question

1.46

 conundrum

1.46

 вопрос

1.45

結論

1.44

 excuses

1.40

Activations Density 0.199%