INDEX

Explanations

analysis of

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

»

-0.07

heits

-0.07

ieder

-0.07

inosaur

-0.07

ftar

-0.07

ager

-0.07

ALLEL

-0.07

å¿«æį·

-0.07

_QMARK

-0.07

ocator

-0.07

POSITIVE LOGITS

åĪĨæŀĲ

0.10

åĪĨæŀĲçļĦ

0.10

åĪĨæŀĲäºĨ

0.10

 analysis

0.10

 ë¶ĦìĦĿ

0.09

åĪĨæŀĲä¸Ģä¸ĭ

0.09

 analyze

0.09

 analyzing

0.08

 analyzes

0.08

åĴĮåĪĨæŀĲ

0.08

Activations Density 0.148%