INDEX

Explanations

adding supplementary information

The neuron strongly activates on adverbs and adjectival modifiers that express degree, frequency, or modality (e.g. “still,” “often,” “also,” “highly”).

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

 continúas

-2.03

'},

-1.94

"},

-1.87

 zijne

-1.81

犼

-1.78

 hvem

-1.68

嗞

-1.61

okra

-1.60

"));

-1.59

 maleta

-1.56

POSITIVE LOGITS

The

2.16

、

1.73

return

1.70

very

1.67

<em>

1.64

See

1.64

even

1.63

You

1.61

1.59

Activations Density 0.069%