INDEX

Explanations

comments and code markers including /*

This neuron fires on numeric digit tokens (i.e. standalone numbers).

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

and

-1.38

 santai

-1.31

chaise

-1.30

}*/

-1.24

 веке

-1.23

//}

-1.22



-1.20

%;

-1.20

६

-1.20

%">

-1.18

POSITIVE LOGITS

では

1.38

<0xA3>

1.29

は

1.23

や

1.22

ですが

1.22

と

1.21

が

1.17

 hilfiger

1.16

Instead

1.15

 instead

1.13

Activations Density 0.000%