INDEX

Explanations

Russian, Polish, and English prepositions/pronouns

neural networks inspired by brains

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 threatening

0.43

 Burgess

0.41

 ongoing

0.40

 accommodating

0.36

 पड

0.36

šti

0.35

抹

0.35

 Burg

0.35

ところに

0.35

となっています

0.34

POSITIVE LOGITS

 тобой

1.71

 ними

1.67

 ним

1.59

 вами

1.54

 этими

1.45

 нами

1.45

 мной

1.43

 nimi

1.42

 этим

1.38

 بهم

1.38

Activations Density 0.006%