INDEX

Explanations

variations of the word "something" with various levels of emphasis

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

isz

-0.07

uthor

-0.07

roll

-0.06

ÑĪÐ¾Ð²

-0.06

nou

-0.06

pis

-0.06

immer

-0.06

AFX

-0.06

isl

-0.06

 trait

-0.06

POSITIVE LOGITS

 like

0.16

 something

0.11

 likes

0.11

 along

0.10

 Like

0.10

_like

0.09

steen

0.09

 seperti

0.09

 LIKE

0.09

Something

0.09

Activations Density 0.005%