INDEX

Explanations

part of

The neuron detects “part of the” markers indicating that the text is identifying a document or article as belonging to a larger series or section.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ၜ

-1.61

 szpital

-1.42

 финан

-1.31

 bicicleta

-1.29

 prodej

-1.29

 Comprometido

-1.27

ⅾ

-1.25

mathrm

-1.25

Ԁ

-1.24

 lihat

-1.23

POSITIVE LOGITS

of

1.98

1.81

1.36

1.30

 упаковка

1.28

We

1.27

previously

1.23

He

1.23

Activations Density 0.104%