INDEX

Explanations

tape, grant, verification

sets x reps

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

ånd

0.48

 χώρα

0.46

 후보

0.46

agis

0.45

žio

0.45

 ಹಾಗ

0.45

hesu

0.45

jor

0.44

yě

0.44

iliar

0.44

POSITIVE LOGITS

ەند

0.51

Excuse

0.48

Walking

0.46

Students

0.45

Holly

0.45

Discrete

0.45

Character

0.45

 identidad

0.45

ِ

0.44

邋

0.44

Activations Density 0.004%