INDEX

Explanations

programming keywords

This neuron activates strongly on code modifiers and access keywords (e.g., public, private, static, readonly, async).

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

 incrí

-1.39

 phénom

-1.34

consejos

-1.31

 unicórnio

-1.29

 tuvimos

-1.29

再一次

-1.29

 éton

-1.29

了自己

-1.28

 distinto

-1.27

ambiance

-1.27

POSITIVE LOGITS

</h2>

1.57

at

1.52

to

1.52



1.39

 first

1.35

)))),

1.32

باح

1.32

 nehme

1.30

1.28

pherals

1.27

Activations Density 0.002%