INDEX

Explanations

code structure and attributes

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

é¡¾

-0.27

’s

-0.26

åħ¨æ°ĳ

-0.26

 contr

-0.25

“

-0.25

 beloved

-0.25

’

-0.25

 overnight

-0.24

 liter

-0.24

“

-0.23

POSITIVE LOGITS

Ø¥Øµ

0.30

Ø£ÙĩØ¯

0.30

Ø§Ø³ØªÙĤÙĦ

0.28

Ø£ÙĪØ¶Ø§Ø¹

0.28

ÙħÙĨØ§Ø·

0.28

ÙħÙģØ§ÙĪØ¶Ø§Øª

0.26

ÙħØ³Ø£ÙĦ

0.26

Ø¥ÙĨØ´

0.26

ä¿İ

0.25

Ø¥Ø¨Ø¯

0.25

Activations Density 0.037%