INDEX

Explanations

presenting an opinion or argument

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

 meestal

-1.25

 waarschijnlijk

-1.17

定価

-1.07

時々

-1.01

どれも

-0.98

ările

-0.98

 praticamente

-0.97

 nobody

-0.97

 ajutor

-0.96

 anybody

-0.94

POSITIVE LOGITS

 argue

1.52

 believe

1.52

認為

1.46

 view

1.42

认为

1.41

 suggests

1.35

 believes

1.35

may

1.24

 even

1.23

 suggest

1.20

Activations Density 0.054%