INDEX

Explanations

adding more descriptive text

The neuron strongly activates on the word “put,” i.e. instances of the verb “put.”

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

ที่จะ

1.44

ST

1.29

eeee

1.28

檛

1.27

ov

1.20

爷

1.20

uar

1.18

ból

1.17

and

1.16

가를

1.14

POSITIVE LOGITS

о

1.30

 onus

1.23

 énfasis

1.21

 forth

1.14

\})

1.09

 emphasis

1.05

ні

1.02

ран

1.02

 firmly

1.02

 undue

1.02

Activations Density 0.055%