INDEX

Explanations

as a

The neuron fires on linking words and verbs that introduce definitions or characterizations—chiefly “as” and its synonyms (e.g. viewed, considered, treated, modeled).

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

 praticamente

-1.23

幾乎

-1.09

tectable

-1.05

接觸

-1.00

 penso

-0.93

 crucial

-0.93

愛的

-0.93

شنگ

-0.91

 δεν

-0.91

 wouldn

-0.91

POSITIVE LOGITS

1.36

 vuitton

1.02

森

0.98

as

0.96

if

0.96

 reisen

0.96

 femininas

0.94

 bébé

0.92

ответ

0.92

 Վերցված

0.91

Activations Density 0.040%