INDEX

Explanations

words and phrases that convey vividness or strong imagery

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

ppard

-0.08

alet

-0.07

	freopen

-0.07

/people

-0.07

UniqueId

-0.06

holding

-0.06

isma

-0.06

oci

-0.06

åĿ¡

-0.06

Ð²ÑģÑı

-0.06

POSITIVE LOGITS

ly

0.14

ness

0.12

LY

0.09

mente

0.09

íŀĪ

0.08

NESS

0.08

çĦ¶

0.08

 enough

0.08

ingly

0.07

ously

0.07

Activations Density 0.003%