INDEX

Explanations

to less than

The neuron activates on numeric tokens or measurements (numbers, including decimals) within 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

of

-1.70

 from

-1.44

 each

-1.36

 Certainly

-1.34

 Since

-1.34

说过

-1.30

the

-1.27

Since

-1.24

 接口

-1.23

襦

-1.23

POSITIVE LOGITS

Hilsen

1.76

 herhangi

1.54

 felices

1.53

薆

1.43

ሖ

1.42

 entspre

1.38

⟣

1.34

 jakieś

1.34

 desperate

1.31

۷

1.31

Activations Density 0.025%