INDEX

Explanations

resource and resources

The neuron consistently fires on occurrences of the word “resource” or “resources.”

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

 atlet

-1.98

</i>

-1.74

})$}

-1.73

 uniquement

-1.73

)」

-1.71

苾

-1.69



-1.69

騖

-1.68

 caufe

-1.68

aculture

-1.68

POSITIVE LOGITS

2.36

 recursos

2.23

is

2.13

 resources

2.05

1.89

1.80

THE

1.77

 газе

1.75

1.72

Activations Density 0.024%