INDEX

Explanations

most followed by adjectives/adverbs

The neuron fires on the intensifier “most,” i.e. superlative degree adverbs indicating the highest extent.

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

-1.98

 homicidio

-1.70

for

-1.55

 from

-1.52

of

-1.52

 طويلة

-1.51

Through

-1.49

ции

-1.48

 melap

-1.46

报

-1.45

POSITIVE LOGITS

有两

1.58

有两个

1.46

 הכי

1.45

 wszystkich

1.45

 สำหรับ

1.38

 najbardziej

1.38

亮点

1.37

 เล

1.36

色々

1.35

上げた

1.35

Activations Density 0.017%