INDEX

Explanations

scenarios and BHP

The neuron strongly fires on tokens that are part of formal titles or headings—especially book, article, or exhibition titles in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

1.92

：

1.61

:...

1.26

:\

1.25

():

1.23

!:

1.22

:',

1.20

.:

1.18

:</

1.15

POSITIVE LOGITS

 برش

0.67

 bizi

0.64

 Rakyat

0.64

jší

0.61

 berbagi

0.61

ΤΑ

0.60

емся

0.58

تیب

0.58

irmek

0.58

czeń

0.57

Activations Density 0.095%