INDEX

Explanations

saving throws and protection

The neuron activates on numeric values and quantitative statistics (e.g. game statistics, measurements, and other numbers).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

to

-1.65

 feminina

-1.38

кладки

-1.34

</b>

-1.29

liziert

-1.28

quelize

-1.26

𝗚

-1.26

 amarela

-1.25

})}

-1.25

ですが

-1.25

POSITIVE LOGITS

från

1.41

following

1.36

 here

1.31

 nessun

1.30

 saving

1.28

lê

1.27

 read

1.23

können

1.22

 protection

1.20

有不

1.20

Activations Density 0.016%