INDEX

Explanations

speaks for itself

The neuron fires on self-referential or meta-linguistic terms—words like “itself,” “become,” “actual,” or “coming”—that point back to or emphasize the discourse itself.

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

 bemer

-1.46

 ainda

-1.36

就行

-1.33

 котором

-1.32

 relevante

-1.32

Pherson

-1.27

出来ました

-1.26

それぞれの

-1.24

 condição

-1.23

してない

-1.23

POSITIVE LOGITS

on

1.63

so

1.61

 måned

1.23

 bahkan

1.23

 This

1.23

 instead

1.22

>),

1.21

]!

1.20

 this

1.20

?>

1.18

Activations Density 0.170%