INDEX

Explanations

numerical scores and outcomes in competitive scenarios

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

gom

-0.08

aran

-0.08

odox

-0.07

osy

-0.07

forge

-0.07

odos

-0.07

 Watts

-0.07

oÄŁ

-0.07

hud

-0.07

dic

-0.07

POSITIVE LOGITS

/-

0.08

âĢ³

0.07

iyon

0.06

ugen

0.06

ideos

0.06

issa

0.06

iken

0.05

 prev

0.05

 final

0.05

.ua

0.05

Activations Density 0.014%