INDEX

Explanations

files, individuals, cells, parties

The neuron fires on key nouns denoting the primary entity or object under manipulation or discussion (e.g. element, cells, files, user, population, vessel, parties).

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

but

-1.55

if

-1.44

 kalau

-1.29

 tett

-1.26

kloped

-1.24

๔

-1.21

↵↵↵

-1.21

actose

-1.19

 tercio

-1.19

-1.16

POSITIVE LOGITS

 financière

1.45

umumkan

1.38

1.37

前面的

1.37

 irgendwo

1.32

 Sehingga

1.30

 først

1.29

淠

1.28

 jemand

1.27

僦

1.27

Activations Density 0.177%