INDEX

Explanations

explaining or confusing

explains or confuses

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

 sérieux

0.46

 unsus

0.44

 victimes

0.42

בודה

0.41

 amplifiers

0.40

ثیر

0.39

 poważ

0.38

 hommes

0.38

အား

0.38

ഗം

0.38

POSITIVE LOGITS

explained

0.67

 confusing

0.62

 explained

0.55

 Explained

0.54

complicated

0.54

Explained

0.54

Explain

0.52

 เหมือน

0.51

也就是说

0.51

 விளக்க

0.50

Activations Density 0.048%