INDEX

Explanations

going through a process or time

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

åĬŁ

-0.10

 dipl

-0.10

 extrapol

-0.10

 extravag

-0.10

 Intercept

-0.09

 rever

-0.09

etch

-0.08

isting

-0.08

echa

-0.08

POSITIVE LOGITS

 ordeal

0.13

 phases

0.12

 hoops

0.11

ordes

0.11

 hell

0.10

 hardships

0.10

 ages

0.10

 heaps

0.10

 Ages

0.10

 stages

0.09

Activations Density 0.062%