INDEX

Explanations

time-related abbreviations indicating morning and evening hours

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

ģµ

-0.06

imenti

-0.06

ç½²

-0.06

unny

-0.06

 small

-0.06

gram

-0.05

 Pope

-0.05

Tap

-0.05

erson

-0.05

and

-0.05

POSITIVE LOGITS

.scalablytyped

0.08

isphere

0.07

 eskort

0.07

ahn

0.07

 aktivit

0.07

/stdc

0.07

Ont

0.07

ÅĻÃŃz

0.07

atform

0.07

Activations Density 0.003%