INDEX

Explanations

confirmation actions or results

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 honn

-0.93

 célè

-0.88

morgen

-0.87

ához

-0.86

 aquellos

-0.85

sistant

-0.85

searchText

-0.85

LoginActivity

-0.84

 melk

-0.84

tront

-0.83

POSITIVE LOGITS

 confirmation

2.66

 thank

2.45

 confirm

2.30

 Thank

2.14

Confirmation

2.06

 success

2.02

Thank

1.97

 final

1.97

 summary

1.96

 Confirmation

1.94

Activations Density 0.034%