INDEX

Explanations

the followed by nouns

The neuron strongly responds to floating‐point numeric literals (numbers containing a decimal point) in the 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

 menghasilkan

-1.46

 břez

-1.46

 októ

-1.46

させて頂

-1.44

 even

-1.39

᎑

-1.39

 schnelle

-1.39

χρι

-1.39

蕺

-1.38

 asig

-1.37

POSITIVE LOGITS

</strong>

1.27

}{

1.25

},

1.25

—

1.23

name

1.21

about

1.20

 solches

1.20

});

1.20

Activations Density 0.060%