INDEX

Explanations

foreign languages

The neuron is essentially inactive (“dead”) and does not detect any pattern in the input.

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

mathrm

-1.59

℉

-1.47

 geschikt

-1.36

ᵴ

-1.33

一个小

-1.30

出来的

-1.23

 färg

-1.23

 peruano

-1.23

玙

-1.21

€/

-1.19

POSITIVE LOGITS

skjaer

1.63

 даже

1.55

 nová

1.53

直到

1.41

 hatten

1.41

rira

1.40

 также

1.38

 damska

1.37

oureuse

1.36

koľvek

1.35

Activations Density 0.001%