INDEX

Explanations

configuration changes

The neuron activates strongly on numeric values and quantitative tokens (e.g. numbers, measurement units, percentages) in the text.

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

addView

0.64

LayoutParams

0.63

Charlotte

0.60

LayoutInflater

0.59

Anch

0.59

getChild

0.58

anch

0.57

JSONException

0.57

美好的

0.55

findViewById

0.55

POSITIVE LOGITS

 transmutation

0.51

 ит

0.48

 Artículo

0.48

원이

0.47

 gating

0.47

 recombin

0.47

cop

0.47

 turbo

0.47

olica

0.47

溼

0.47

Activations Density 0.003%