INDEX

Explanations

phrases indicating contrasting ideas or conditions

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

hani

-0.07

ahun

-0.07

tvrt

-0.07

iren

-0.06

 purely

-0.06

 Pretty

-0.06

áº¡n

-0.06

RACT

-0.06

essional

-0.06

Pretty

-0.05

POSITIVE LOGITS

 especially

0.27

 especialmente

0.24

 particularly

0.23

especially

0.23

å°¤

0.22

particularly

0.19

 besonders

0.19

 Especially

0.18

çī¹åĪ«

0.18

 Ã¶zellikle

0.17

Activations Density 0.139%