INDEX

Explanations

listing examples or elaborations

The neuron is sensitive to capitalized tokens—proper names, titles, acronyms, and other words beginning with an uppercase letter.

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

but

-2.34

 with

-1.54

 that

-1.39

for

-1.31

to

-1.29

 because

-1.25

šo

-1.24

 чтобы

-1.23

 poiché

-1.18

が

-1.16

POSITIVE LOGITS

なんですよ

1.57

なんですね

1.50

為に

1.45

 linkovi

1.32

 manquer

1.32

 useAppContext

1.30

 willen

1.29

艄

1.28

ohne

1.28

besondere

1.27

Activations Density 0.243%