INDEX

Explanations

public declarations and identifiers

The neuron fires on numeric literal tokens (the actual 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

With

-1.33

！（

-1.27

。

-1.23

!!!!!

-1.22

!!!!

-1.22

-1.21

 температуре

-1.18

 新品

-1.17

kozások

-1.17

！(

-1.16

POSITIVE LOGITS

 same

1.70

 then

1.55

or

1.46

 distintos

1.42

 considerados

1.41

 oorspron

1.40

 aislados

1.40

 therein

1.39

 jego

1.39

 azért

1.37

Activations Density 0.003%