INDEX

Explanations

loan or help

The neuron strongly activates on numeric tokens that include a decimal point (i.e. floating‐point numbers).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 dissidents

0.94

 backers

0.88

 deems

0.83

alate

0.80

 whip

0.80

bedarf

0.80

кет

0.79

 Künst

0.79

 denounced

0.79

 commended

0.78

POSITIVE LOGITS

OpenGL

0.83

0.80

Search

0.74

↵

0.74

时光

0.73

IndexOf

0.72

SHIFT

0.71

게

0.70

Header

0.69

Surface

0.69

Activations Density 0.000%