INDEX

Explanations

end punctuation

The neuron activates specifically on four-digit numeric tokens.

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

🤓

-1.47

🤫

-1.30

が

-1.30

😤

-1.27

🤭

-1.25

😋

-1.24

 quartos

-1.24

 australis

-1.23

💗

-1.23

🥺

-1.23

POSITIVE LOGITS

って何

1.61

久し

1.60

 humedad

1.59

牴

1.55

 delicada

1.51

はコチラ

1.44

 colorear

1.42

はこちらから

1.41

念头

1.41

縝

1.41

Activations Density 0.219%