INDEX

Explanations

modal verbs followed by outcomes

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 hugs

1.46

 wilds

1.41

 borrows

1.35

 overshoot

1.30

 economies

1.25

 resembles

1.25

 multiplications

1.25

 COMPO

1.24

 Nearest

1.24

äle

1.23

POSITIVE LOGITS

 pression

1.63

まま

1.62

圧力

1.60

lett

1.59

姿勢

1.56

dne

1.52

חר

1.51

 pressione

1.49

サイン

1.49

lete

1.47

Activations Density 0.041%