INDEX

Explanations

HTML img, insert statements, options

The neuron detects tokens containing numbers (numeric literals or digit‐filled strings) in code or markup.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 both

-1.62

 their

-1.34

 that

-1.30

 perdido

-1.28

 estampado

-1.27

 armado

-1.25

how

-1.22

 either

-1.18

 vermelhas

-1.16

≳

-1.15

POSITIVE LOGITS

 incluso

1.41

amd

1.12

ul

1.02

0.99

tab

0.96

↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.93

說道

0.93

StatusCodes

0.93

↵↵↵↵↵↵↵↵↵↵↵↵↵

0.92

ron

0.91

Activations Density 0.032%