INDEX

Explanations

defining primary keys

This neuron activates on programming‐style syntax tokens—SQL keywords, comment markers (“--”, “#”), ellipses, semicolons, etc.—i.e. signs of code structure.

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

 ospiti

-1.49

utusan

-1.44

its

-1.38

-1.36

endő

-1.33

avano

-1.32

éndole

-1.31

éndolo

-1.29

 transacción

-1.28

urati

-1.27

POSITIVE LOGITS

are

1.38

1.36

 dominio

1.35

などが

1.33

for

1.33

1.28

</h1>

1.27

 поэтому

1.23

Activations Density 0.062%