INDEX

Explanations

I am/have/cannot

The neuron activates strongly on personal pronouns—especially first-person (“I,” “we”) and also “you.”

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

ли

0.98

0.81

0.74

ためには

0.71

ifs

0.70

ي

0.68

 amerikan

0.66

li

0.64

ni

0.64

 when

0.62

POSITIVE LOGITS

 نے

0.57

getWindow

0.56

сное

0.56

 중심으로

0.55

aconit

0.55

varage

0.55

ровать

0.54

casualties

0.53

の水

0.53

的水

0.52

Activations Density 0.541%