INDEX

Explanations

code syntax

The neuron activates on numeric literal tokens (e.g. standalone numbers like 72, 67.5, 45, etc.).

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

 教程

-1.53

(!_

-1.52

 Seperti

-1.48

 搭配

-1.47

 étrang

-1.34

where

-1.30

 洗濯

-1.29

 搞笑

-1.27

 Âge

-1.27

 Misalnya

-1.25

POSITIVE LOGITS

 name

1.60

为何

1.54

 शुर

1.43

岀

1.42

Læs

1.37

 juos

1.37

 οποίο

1.36

何时

1.35

独特的

1.34

 daugiau

1.33

Activations Density 0.018%