INDEX

Explanations

conditional phrases involving the word "you."

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

jerne

-0.07

ond

-0.07

ifu

-0.07

hek

-0.07

Ì£

-0.07

emailer

-0.06

ignon

-0.06

 Hogan

-0.06

 Parkway

-0.06

IDL

-0.06

POSITIVE LOGITS

 prefer

0.08

acific

0.07

 Prefer

0.07

 preference

0.07

stÃŃ

0.07

 preferred

0.07

 more

0.07

 prefers

0.06

 Ð¿ÑĢÐµÐ´Ð¿Ð¾Ñĩ

0.06

XL

0.06

Activations Density 0.017%