INDEX

Explanations

or

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

ä¸įä½Ĩ

-0.08

ä¹Łæĺ¯è¦ģ

-0.08

iak

-0.07

ä¸įå±ŀäºİ

-0.07

resh

-0.07

ersp

-0.06

COMPARE

-0.06

asih

-0.06

Ã©r

-0.06

HECK

-0.06

POSITIVE LOGITS

 equival

0.09

ients

0.09

Ð¾ÑģÑĤÐ¾

0.07

ally

0.07

inio

0.07

plevel

0.07

 alternatively

0.07

oran

0.07

ä¹ĭä¸º

0.07

rawat

0.07

Activations Density 0.037%