INDEX

Explanations

file format documents

The neuron activates on Chinese-language tokens (e.g. “文件”), i.e. it detects Chinese script in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits



-2.89

ÁRIO

-2.47

IConfiguration

-2.30

’

-2.25

 correctes

-2.20



-2.20

бора

-2.16

 ‘’

-2.16

AN

-2.13

’,

-2.11

POSITIVE LOGITS

2.50

 spécialisé

2.47

 didalam

2.41

所で

2.28

他有

2.20

最初の

2.19

殽

2.16

2.14

are

2.13

不该

2.13

Activations Density 0.001%