INDEX

Explanations

numbers and dates

This neuron detects occurrences of personal pronouns (especially “you” and “it”) and associated contractions or apostrophes, signaling direct address or informal tone.

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

że

-1.53

시간

-1.41

！(

-1.38

<>());

-1.36

«

-1.34

 getItemId

-1.29

že

-1.28

newblock

-1.28

_;

-1.27

áže

-1.27

POSITIVE LOGITS

 both

1.58

 feuer

1.53

 blumen

1.45

 such

1.40

 these

1.36

</u>

1.35

 xícara

1.34

 stanga

1.33

 seite

1.32

 rhein

1.31

Activations Density 0.137%