INDEX

Explanations

UK politics and charts

The neuron activates on numeric tokens (e.g. counts, IDs, measurements, years) whenever numbers appear in 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

اشق

-1.05

fú

-0.87

 causado

-0.83

 Grims

-0.79

养成

-0.77

員工

-0.77

CUT

-0.76

fadeOut

-0.75

hofer

-0.73

 ית

-0.73

POSITIVE LOGITS

孩

0.86

roient

0.84

igraphy

0.81

呀

0.80

afir

0.79

RESOLUTION

0.76

UK

0.76

jojo

0.75

slowly

0.75

τού

0.74

Activations Density 0.006%