INDEX

Explanations

phrases indicating preparation or buildup to an event

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

 place

-0.06

Fut

-0.06

ongan

-0.06

thr

-0.06

iras

-0.06

izable

-0.06

_Impl

-0.06

Fon

-0.06

ken

-0.05

 halves

-0.05

POSITIVE LOGITS

iales

0.07

ÑħÐ¸

0.07

alsex

0.07

/post

0.07

erra

0.07

uxtap

0.07

Sibling

0.07

igans

0.07

Unnamed

0.06

WebResponse

0.06

Activations Density 0.007%