INDEX

Explanations

since

The neuron strongly activates on words that denote actions or events (verbs in various tenses and forms).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 when

-1.52

atoto

-1.27

if

-1.26

 complementos

-1.23

ながらも

-1.22

 języka

-1.20

 razie

-1.19

 powied

-1.17

 felicitación

-1.17

StatusCodes

-1.16

POSITIVE LOGITS

 since

1.44

了出去

1.41

ђ

1.33

1.32

Since

1.29

herin

1.27

since

1.24

//

1.23

一是

1.22

anas

1.20

Activations Density 0.030%