INDEX

Explanations

pronoun subjects performing actions

The neuron strongly activates on third-person singular personal pronouns (e.g., he, she, him, her, his).

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

of

-2.59

 jego

-1.54

från

-1.49

 Presidencia

-1.49

БЕ

-1.49

diyah

-1.47

țiilor

-1.42

 αυτή

-1.42

 från

-1.41

 cocok

-1.39

POSITIVE LOGITS

1.59

 करती

1.45

1.40

if

1.34

1.31

を行いました

1.27

ke

1.26

了一些

1.26

ers

1.24

ല

1.23

Activations Density 0.046%