INDEX

Explanations

you do/does/did

The neuron strongly activates on the auxiliary verbs “do” and “does.”

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

-1.57

筈

-1.34

 antemano

-1.33

 anfitrión

-1.33

際に

-1.29

えっ

-1.28

あっという

-1.26

 continually

-1.26

ふと

-1.22

sero

-1.20

POSITIVE LOGITS

⼈

1.33

 when

1.28

妖精

1.23

姪

1.20

ínica

1.19

garek

1.19

 penge

1.16

͏

1.16

捺

1.14

 pueda

1.13

Activations Density 0.018%