INDEX

Explanations

a bit, a little, a few

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

jak

-0.10

-0.09

deb

-0.09

Fin

-0.09

-volume

-0.09

LLU

-0.08

 advances

-0.08

 upfront

-0.08

 Featured

-0.08

 finely

-0.08

POSITIVE LOGITS

bit

0.12

 despre

0.11

 fraction

0.11

 notch

0.10

few

0.10

ryo

0.09

 medium

0.09

bit

0.09

 filtr

0.09

 fractions

0.09

Activations Density 0.008%