INDEX

Explanations

regress, confusion, doubt

ratings and durations

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 ওয়া

0.51

 シェ

0.51

 பராம

0.50

 жана

0.48

 sécr

0.48

্নি

0.46

ρηση

0.46

্মান

0.46

}]

0.45

 исклю

0.45

POSITIVE LOGITS

 repent

0.56

 confuse

0.47

 Vign

0.47

 despair

0.46

tin

0.45

 confusion

0.44

 negativo

0.43

 regress

0.43

 timid

0.43

 repentance

0.42

Activations Density 0.001%