INDEX

Explanations

alpha, Alpha, alfa, ALPHA

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ute

-0.09

ycz

-0.09

 halfway

-0.09

brick

-0.09

 Middle

-0.09

 Quick

-0.09

UTE

-0.08

oto

-0.08

il

-0.08

 fines

-0.08

POSITIVE LOGITS

 Alpha

0.15

 ALPHA

0.14

.Alpha

0.14

,alpha

0.14

Beta

0.13

-beta

0.13

alfa

0.13

(alpha

0.13

alph

0.13

 Omega

0.12

Activations Density 0.086%