INDEX

Explanations

This neuron activates on numeric literals—especially the integer values in `#define` macro lines.

macro definitions

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

 함께

-1.40

了什么

-1.31

 those

-1.26

 berbagai

-1.24

<bos>

-1.24

 other

-1.22

 dlaczego

-1.21

ต่างๆ

-1.20

 whatnot

-1.20

 تلك

-1.17

POSITIVE LOGITS

 februari

1.36

 spørgsmål

1.33

 mina

1.33

1.31

 soos

1.29

<>

1.27

“

1.27

 januari

1.27

𝑥

1.24

1.22

Activations Density 0.013%