INDEX

Explanations

bag

This neuron detects occurrences of the substring “bag” (i.e. the word “bag” and its compounds).

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

—

-3.05

 哪些

-2.83

臯

-2.78

摃

-2.73

愎

-2.69

while

-2.58

from

-2.55

<em>

-2.44

↵

-2.44

》，

-2.38

POSITIVE LOGITS

 pady

2.53

 сказала

2.48

इटम

2.47

kingen

2.45

aikan

2.36

 mencegah

2.36

&#

2.34

 indican

2.33

iding

2.27

 padx

2.23

Activations Density 0.008%