INDEX

Explanations

code statements after colon

The neuron flags the punctuation tokens (the “?” and “:”) used in conditional (ternary) expressions.

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

it

-1.59

’

-1.49

when

-1.36

↵

-1.32

 καλύτε

-1.17

 лучше

-1.16

 meditative

-1.16

 while

-1.13

 verbess

-1.11

POSITIVE LOGITS

潾

1.80

 naturligt

1.60

 nesten

1.50

irmanship

1.47

 fredag

1.45

 Diputados

1.41

 secrétaire

1.39

 onsdag

1.38



1.38

 virkelig

1.38

Activations Density 0.005%