INDEX

Explanations

color values, hex codes

The neuron activates on numeric literals containing a decimal point (i.e. floating-point numbers) in code.

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

ことになる

-1.35

it

-1.34

 também

-1.28

ใคร

-1.24

 también

-1.23

をご覧ください

-1.23

 degradability

-1.21

 rénovation

-1.20

 Then

-1.19

었습니다

-1.19

POSITIVE LOGITS

⅖

1.53

 only

1.38

ÇÃO

1.30

="",

1.30

ZIONE

1.28

 other

1.27

 telah

1.25

⅗

1.21

if

1.17

 embarazadas

1.16

Activations Density 0.015%