INDEX

Explanations

code comments and programming languages

The neuron activates on standalone numeric tokens (sequences of digits).

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

が

-1.45

～！

-1.28

 Græ

-1.24

最近は

-1.23

<bos>

-1.21

を

-1.18

 ваз

-1.16

は

-1.16

に

-1.13

PLICATIONS

-1.12

POSITIVE LOGITS

為に

1.41

 verschillende

1.34

and

1.31

感じが

1.31

場所で

1.22

 diversos

1.20

 diferentes

1.20

に行って

1.20

 autén

1.19

の方に

1.16

Activations Density 0.063%