INDEX

Explanations

To be followed by adjective

phrases and their follow-ups

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 abhid

0.40

 violates

0.38

 ግራም

0.38

 vutta

0.37

 blanca

0.37

coerce

0.37

 rougeâtre

0.36

 rnorm

0.36

 sayam

0.36

ጲ

0.36

POSITIVE LOGITS

and

0.42

0.40

 Therapy

0.40

0.38

 Youth

0.38

 Partnership

0.38

 Management

0.37

 Memorial

0.37

 Advisory

0.37

Activations Density 0.028%