INDEX

Explanations

describing current states or past conditions

The neuron strongly activates on copular “to be” and related linking verbs (e.g. am, is, are, be).

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

льники

-1.25

 libri

-1.24

 apartament

-1.21

には

-1.20

 departament

-1.20

توضیحات

-1.19

 opin

-1.15

だから

-1.14

 also

-1.13

IndexOutOfBounds

-1.13

POSITIVE LOGITS

ところに

1.47

為に

1.44

感じで

1.42

来週

1.39

向けの

1.33

って何

1.30

の方に

1.30

うわ

1.26

わから

1.24

 któ

1.19

Activations Density 0.096%