INDEX

Explanations

system and its variants

The main thing this neuron does is detect the word “system” (and its variants like “systems,” “system’s,” etc.) in technical contexts.

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

rows

0.62

igte

0.61

ペース

0.60

 browse

0.58

ilia

0.57

シート

0.57

 streak

0.56

時点

0.56

からです

0.56

Kej

0.56

POSITIVE LOGITS

 система

1.38

 sistema

1.30

 системи

1.26

 sistemi

1.25

 systemu

1.24

 системы

1.22

 système

1.20

 system

1.20

系统

1.20

 систему

1.20

Activations Density 0.414%