INDEX

Explanations

Mr. [name]

The neuron chiefly activates on the honorific “Mr.” 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

-2.03

jusqu

-1.78

</em>

-1.73

 haupts

-1.73

 delegación

-1.72

 tiroirs

-1.69

 comisión

-1.66

 Kartoffeln

-1.66

。

-1.66

 gebraucht

-1.61

POSITIVE LOGITS

 What

2.02

 While

2.00

 Though

1.97

 However

1.90

 Even

1.87

</h1>

1.83

 different

1.82

 Through

1.81

 From

1.79

 Without

1.79

Activations Density 0.024%