INDEX

Explanations

JavaScript and React code

The neuron fires strongly on class-declaration and inheritance keywords and names (e.g. “class … extends …”) 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

にしました

-1.07

だけです

-1.03

描いて

-1.02

-1.01

 relationship

-0.99

完成です

-0.98

cellor

-0.97

 preocupação

-0.96

ड़ा

-0.96

 пальто

-0.96

POSITIVE LOGITS

␮

1.27

 asfal

1.24

-------

1.23

 tortas

1.15

 tarta

1.15

 sopa

1.13

---------

1.12

 bombas

1.11

 enctype

1.11

 lagar

1.09

Activations Density 0.017%