INDEX

Explanations

house of

The neuron strongly activates on occurrences of the word “house.”

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

躄

-1.56

ation

-1.51

vedades

-1.49

wwww

-1.48

lizes

-1.46

觧

-1.42

椛

-1.41

薷

-1.41

になるので

-1.41

ressee

-1.41

POSITIVE LOGITS

of

1.98

出了一

1.73

 equally

1.62

 also

1.61

not

1.48

 abit

1.30

水墨

1.30

Ingredienti

1.29

春秋

1.27

no

1.25

Activations Density 0.012%