INDEX

Explanations

scientific research

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

 REUTERS

-0.22

umberland

-0.21

ynchronously

-0.21

oggler

-0.21

.innerHeight

-0.20

allis

-0.20

vertisement

-0.20

.AutoScaleMode

-0.20

 ])->

-0.20

,");↵

-0.20

POSITIVE LOGITS

çĽ¸å¯¹äºİ

0.28

 regarding

0.28

 itself

0.27

 relative

0.26

wr

0.25

å¯¹ä¼ģä¸ļ

0.25

vis

0.24

à¹ĥà¸Ļà¸ģà¸²à¸£

0.24

 against

0.24

å¯¹åħ¶

0.24

Activations Density 0.765%