INDEX

Explanations

the word 'one' and phrases that emphasize singularity or prominence

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

-0.07

axe

-0.07

-UA

-0.06

hlas

-0.06

wend

-0.06

æ·¡

-0.06

AREN

-0.06

ãģ®åŃĲ

-0.06

 shades

-0.06

GES

-0.06

POSITIVE LOGITS

 such

0.11

åħ¶ä¸Ń

0.10

such

0.09

 Such

0.08

ãģĿãģĨ

0.08

Such

0.07

 those

0.07

 ê·¸ëŁ¬

0.07

 example

0.07

 among

0.07

Activations Density 0.015%