INDEX

Explanations

be sure, ready, aware, alert

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

BG

-0.10

kur

-0.09

anders

-0.08

 Damien

-0.08

 troublesome

-0.08

osti

-0.08

DT

-0.08

 gladly

-0.08

 redis

-0.08

 balk

-0.08

POSITIVE LOGITS

 careful

0.26

 prepared

0.23

 aware

0.22

 Prepared

0.19

 sure

0.19

 mindful

0.17

 vigilant

0.17

prepared

0.17

 vers

0.17

 cautious

0.16

Activations Density 0.070%