INDEX

Explanations

The neuron fires on sequences of uppercase initials or abbreviated capital-letter codes (e.g., two-letter name initials or acronyms).

people's initials

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

to

-1.27

価格は

-1.27

 first

-1.27

義務

-1.20

 then

-1.20

was

-1.20

 actually

-1.18

結果は

-1.18

使い方

-1.16

内容は

-1.16

POSITIVE LOGITS

 THOSE

1.38

の方に

1.30

ſelf

1.30

🔙

1.23

事が

1.23

 provoquer

1.22

 DÍA

1.21

 aqueles

1.20

 francia

1.20

Activations Density 0.024%