INDEX

Explanations

only followed by a noun or verb

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 whatsoever

-0.12

-any

-0.10

semblies

-0.09

 kitt

-0.09

 Fitzgerald

-0.09

 carving

-0.09

/extensions

-0.09

ucer

-0.09

INC

-0.09

 McCorm

-0.09

POSITIVE LOGITS

 only

0.14

 ÑĤÐ¾Ð»ÑĮÐºÐ¾

0.12

 lasted

0.11

 reason

0.11

only

0.10

 occasional

0.10

 reply

0.10

 briefly

0.10

Only

0.09

 occasionally

0.09

Activations Density 0.002%

No Known Activations