INDEX

Explanations

asking questions

The neuron fires on tokens that signal someone asking a question (e.g. “asked,” “why,” “if”).

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

 ermöglicht

-1.52

</h2>

-1.49

↵

-1.41

 ähnlichen

-1.40

-1.37

 extremely

-1.33

。

-1.33

 also

-1.33

-1.32

jn

-1.31

POSITIVE LOGITS

liputi

1.42

ispiele

1.28

 there

1.26

 Vitamina

1.24

ίη

1.22

 çantası

1.20

 chère

1.19

ṿ

1.19

Über

1.18

 "..\..\

1.18

Activations Density 0.017%