INDEX

Explanations

strong evidence or understanding

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

imetsu

-1.67

 安心

-1.49

We

-1.48

 such

-1.45

</strong>

-1.42

urança

-1.37

vestidos

-1.37

 personen

-1.36

brochen

-1.33

 Privacidad

-1.31

POSITIVE LOGITS

strong

2.25

 strong

2.08

Strong

1.81

 💪

1.67

 starken

1.67

 kuat

1.65

STRONG

1.62

 stronger

1.61

 Strong

1.57

强的

1.56

Activations Density 0.054%