INDEX

Explanations

numbers like 900 or 9000

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

𖣘

-2.33

 häng

-2.31



-2.22

咴

-2.20



-2.16

 rotten

-2.14

 喜欢

-2.13

 daring

-2.09

 και

-2.08

 shatter

-2.08

POSITIVE LOGITS

 indeb

2.42

 специфи

2.20

 They

2.19

an

2.17

更何况

2.06

’

2.05

灕

2.02

弢

1.98

 avena

1.95

 agrico

1.94

Activations Density 0.006%