INDEX

Explanations

punctuation and conjunctions indicating lists or connections

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

jin

-0.07

emat

-0.07

RAIN

-0.07

isin

-0.06

allen

-0.06

 Ã©c

-0.06

-:-

-0.06

/watch

-0.06

ntl

-0.06

ÑĪÑĥ

-0.06

POSITIVE LOGITS

 yours

0.06

ering

0.06

 Role

0.06

udoku

0.06

 Tillerson

0.05

artment

0.05

anto

0.05

Ø§Ø¡

0.05

ibold

0.05

 Lang

0.05

Activations Density 0.040%