INDEX

Explanations

expressing gratitude

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

chetto

-0.07

å¼ķäººåħ¥

-0.07

keur

-0.07

èĪĪåĳ³

-0.07

 disastr

-0.07

Assert

-0.07

coni

-0.07

UnderTest

-0.06

_ASSUME

-0.06

(![

-0.06

POSITIVE LOGITS

æĦŁæ¿Ģ

0.16

 gratitude

0.16

æĦŁè°¢

0.16

 Ð±Ð»Ð°Ð³Ð¾Ð´Ð°ÑĢ

0.16

æĦŁè¬Ŀ

0.15

æĦŁæģ©

0.15

 thanking

0.14

 grateful

0.14

 Ð±Ð»Ð°Ð³Ð¾Ð´Ð°ÑĢÐ½Ð¾ÑģÑĤÑĮ

0.14

 thank

0.14

Activations Density 0.195%