INDEX

Explanations

even or example

The neuron fires on first-person discourse markers (like “I mean,” “I’m,” “Hell, I’m,” etc.) that introduce the speaker’s personal commentary or emphasis.

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

 these

-1.09

È

-0.91

 olika

-0.90

 verkligen

-0.89

 theſe

-0.89

Netherlands

-0.86

}_{\

-0.86

 этих

-0.85

 thofe

-0.85

$_{

-0.85

POSITIVE LOGITS

 even

2.91

 bahkan

1.69

even

1.68

 например

1.66

 даже

1.66

甚至

1.64

Even

1.63

 nawet

1.57

 навіть

1.55

 incluso

1.51

Activations Density 0.024%