INDEX

Explanations

numbers followed by punctuation or units

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

久し

-1.42

あぁ

-1.37

えっ

-1.36

 fré

-1.28

総評

-1.27

＝＝＝＝

-1.27

何を

-1.26

我が家の

-1.22

・・・・・・・・

-1.21

 obé

-1.20

POSITIVE LOGITS

が

1.24

↵

1.22

で

1.19

TokenNameDOT

1.18

<bos>

1.13

 “(

1.12

还有

1.11

は

1.09

 mehrere

1.09

Activations Density 0.001%