INDEX

Explanations

explain what

asking to explain

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

 заяви

0.67

 рекоменда

0.63

 победи

0.61

 acknowledge

0.57

ní

0.56

 erkennen

0.56

 взя

0.55

 Север

0.55

ncies

0.54

elä

0.54

POSITIVE LOGITS

 explain

1.03

 Explanation

1.02

Explain

1.02

 توضیح

1.01

 설명

0.99

解释

0.98

解釋

0.97

 explains

0.97

 Explain

0.97

 explanation

0.93

Activations Density 0.156%