INDEX

Explanations

limited, unlimited, new, entire

The neuron fires on tokens that express quantities, amounts, durations, or degrees (e.g. numbers, time spans, and quantifier/adjective intensifiers).

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

歟

-1.50

🛇

-1.38

スマートフォン

-1.37

英寸

-1.33

ῖ

-1.32

 aéro

-1.31

🦼

-1.30

-1.28

 suscit

-1.27

 sayfas

-1.27

POSITIVE LOGITS

 stuff

2.14

 thing

1.80

 things

1.63

but

1.51

 info

1.37

 from

1.34

 people

1.34

 maybe

1.34

 ändå

1.30

now

1.28

Activations Density 0.267%