INDEX

Explanations

and then words

The neuron specifically detects the conjunction "and" in running 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

 овощи

-1.41

 what

-1.22

 našich

-1.21

 barvy

-1.21

 گیاه

-1.21

 творо

-1.21

 ۱۳۸

-1.20

íme

-1.20

 المياه

-1.20

 kalori

-1.19

POSITIVE LOGITS

 said

1.79

ֵּ

1.45

ַּ

1.38

 officials

1.37

ométrique

1.35

ύ

1.34

וּ

1.28

 authorities

1.26

 says

1.25

 there

1.23

Activations Density 0.046%