INDEX

Explanations

variations of the word "best" indicating rankings or evaluations

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

Ä±rak

-0.09

ipers

-0.07

ombine

-0.07

à¸Ńà¸ļ

-0.07

poon

-0.07

idious

-0.07

eyen

-0.07

.Interop

-0.07

PLETED

-0.07

theid

-0.07

POSITIVE LOGITS

 best

0.07

 Karlov

0.07

 Best

0.07

 ever

0.07

 sight

0.06

VS

0.06

bet

0.06

ows

0.06

aux

0.06

-in

0.06

Activations Density 0.012%