INDEX

Explanations

names following roles

names and researchers

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

 دکتر

0.38

জনাব

0.37

 الدكتور

0.35

 प्रोफेसर

0.31

 মিসেস

0.31

 dokter

0.29

 niej

0.29

 ഡോ

0.29

فرم

0.29

 டாக்டர்

0.29

POSITIVE LOGITS

and

0.29

0.28

 attractiveness

0.25

 consolation

0.25

Â

0.24

 connection

0.23

0.22

↵↵

0.22

Ã

0.22

Activations Density 0.044%