INDEX

Explanations

sentence ending punctuation

The neuron fires strongly on descriptive or evaluative words (mostly adjectives and adjectival participles) that express an assessment or judgment.

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

 that

-2.28

 until

-2.06

-1.95

in

-1.95

of

-1.90

 before

-1.87

for

-1.83

for

-1.80

if

-1.71

who

-1.68

POSITIVE LOGITS

 abou

1.82



1.76

Démographie

1.73

 никогда

1.73

Функ

1.73

چون

1.73

CHREIB

1.73

 spécialistes

1.72

AUTOR

1.72

 alltid

1.72

Activations Density 0.023%