INDEX

Explanations

computer names

This neuron flags individual capitalized “name” tokens—e.g. parts of proper nouns, titles, or branded terms.

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

on

-1.86

of

-1.48

 often

-1.41

 stated

-1.30

 first

-1.29

 using

-1.26

 described

-1.26

 expected

-1.23

 used

-1.21

POSITIVE LOGITS

',\

1.59

\",\

1.58

AutoScaleMode

1.38

'];

1.34

特に

1.31

//};

1.31

</em>

1.30

};

1.29

 Yearly

1.28

 numéros

1.28

Activations Density 0.149%