INDEX

Explanations

instances of the word "para" and its variations, indicating a focus on purpose or intention

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

reich

-0.07

ombo

-0.06

do

-0.06

lose

-0.06

haft

-0.06

rossover

-0.06

.adv

-0.06

cribe

-0.06

 ÅĻÃŃj

-0.06

Dou

-0.06

POSITIVE LOGITS

sÃŃ

0.09

tant

0.09

bidden

0.08

anuts

0.08

ãģªãģĦ

0.08

ady

0.08

 sake

0.07

Ã®tre

0.07

Ùĩ

0.07

Activations Density 0.014%