INDEX

Explanations

Rogue, boutique

The neuron selectively fires on the literal string “rogue” wherever it appears 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

re

-0.99

 großer

-0.94

den

-0.84

jenigen

-0.84

 впер

-0.83

 Räume

-0.82

ética

-0.79

 realize

-0.78

图表

-0.77

↵↵

-0.76

POSITIVE LOGITS

like

1.10

likes

0.88

 Viernes

0.87

LIKE

0.87

↝

0.87

 sper

0.85

Rogue

0.85

isha

0.83

 član

0.82

😜

0.82

Activations Density 0.013%