INDEX

Explanations

selection criteria

negative sentiment or inability to act

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Pelican

0.46

 Publ

0.45

 तहसीलदार

0.45

 Peoples

0.45

 designee

0.44

 Mens

0.44

 Churches

0.44

ázky

0.43

橹

0.43

 Shipping

0.42

POSITIVE LOGITS

 deteriorating

0.54

{\

0.54

SR

0.52

Ста

0.52

IEEE

0.52

 worsening

0.52

active

0.50

Const

0.50

Р

0.50

}{\

0.49

Activations Density 0.001%