INDEX

Explanations

secret, precision

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 strict

-1.02

strict

-0.93

 secret

-0.89

 tight

-0.84

 Strict

-0.77

 closely

-0.76

tight

-0.73

 tightly

-0.73

secret

-0.71

 strictly

-0.71

POSITIVE LOGITS

 vayas

0.69

||=

0.66

 Magnetism

0.65

al

0.63

 sumpay

0.60

 Daarna

0.59

 mierda

0.58

 snuff

0.57

 mapStateToProps

0.56

 altars

0.56

Activations Density 0.043%