phrases indicating intention or purpose

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥ¼ãĥĨ

-0.08

tut

-0.07

verte

-0.07

jk

-0.07

 VerfÃ¼gung

-0.07

Ð¾Ð³ÑĢÐ°Ð¼

-0.07

alia

-0.07

ï¸

-0.07

meer

-0.07

 coward

-0.06

POSITIVE LOGITS

ondo

0.07

 persuade

0.06

 excess

0.06

abs

0.06

ATALOG

0.06

awks

0.06

OE

0.05

Activations Density 0.035%