INDEX

Explanations

master or True

The neuron strongly activates on tokens representing numbers or numeric references (e.g. digits, quantities, measurements).

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

AD

0.76

UNIT

0.73

with

0.72

Enc

0.72

Abb

0.72

…).

0.72

Г

0.72

August

0.71

...).

0.71

cond

0.70

POSITIVE LOGITS

өлү

0.86

ыска

0.86

 эволю

0.83

 мэ

0.82

शुदा

0.82

ᡳ

0.81

వాన్ని

0.80

 merkez

0.79

órias

0.79

ным

0.78

Activations Density 0.000%