INDEX

Explanations

quadratic expressions and equations

The main thing this neuron does is detect numeric values or numeric expressions (digits and numbers) in text.

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

</b>

-2.16

to

-1.83

幸好

-1.56

只剩下

-1.48

Iam

-1.44

really

-1.40

 berharap

-1.38

 apunta

-1.34

believe

-1.33

/></

-1.31

POSITIVE LOGITS

</h3>

1.75

1.41

第一个

1.40

 sonriente

1.39

 agujas

1.38

 деца

1.36

mbito

1.30

áveis

1.30

 sixteen

1.30

 یا

1.27

Activations Density 0.079%