INDEX

Explanations

The neuron detects first-person self-referential statements, especially “I” followed by a verb expressing personal experience.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

our

1.38

 आम्हाला

1.24

 আমাদেরকে

1.20

 हमें

1.18

我们的

1.16

 ہمیں

1.16

 bize

1.15

 themselves

1.14

 ours

1.12

我们要

1.12

POSITIVE LOGITS

 myself

2.22

我自己

1.97

 دارم

1.95

 خودم

1.84

 jestem

1.84

 છું

1.80

хожу

1.77

 हूं

1.73

 люблю

1.71

 работаю

1.70

Activations Density 1.646%