INDEX

Explanations

for a specific purpose

The neuron preferentially activates on very common function words (e.g. articles, pronouns, prepositions).

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

 trucos

-1.12

 hjælpe

-1.06

paña

-1.05

 både

-1.05

 more

-1.02

 dormitorios

-1.02

 trámites

-0.96

 oldal

-0.96

ᴎ

-0.91

äjä

-0.91

POSITIVE LOGITS

たった

1.78

เพียง

1.76

 only

1.66

 chỉ

1.63

 einzigen

1.62

 فقط

1.56

 just

1.50

 только

1.48

 merely

1.47

 apenas

1.47

Activations Density 0.074%