INDEX

Explanations

modelforeign language or programming elements

The neuron activates on numeric tokens—strings of digits (including numbers with decimals) such as version numbers, timestamps, and other numeric constants.

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

 jugó

-1.41

dersfield

-1.38

 tiver

-1.37

 while

-1.29

írez

-1.26

 отношению

-1.25

***********/

-1.20

liage

-1.16

 until

-1.16

utdown

-1.16

POSITIVE LOGITS

 they

1.55

很高

1.52

他們

1.46

 these

1.31

他们

1.30

you

1.30

越高

1.27

1.26

 musíte

1.26

 эти

1.25

Activations Density 0.203%