INDEX

Explanations

difficult situations or predicaments

The neuron activates on abstract, often weighty nouns that denote states, circumstances, or existential conditions (e.g., “existence,” “condition,” “plight,” “predicament”).

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

the

-1.95

凩

-1.68

鹮

-1.56

It

-1.51

%),

-1.48

 whom

-1.45

亊

-1.38

 апре

-1.34

nha

-1.34

 attraverso

-1.33

POSITIVE LOGITS

其他

1.41

這些

1.30

or

1.29

ني

1.28

These

1.27

 aktar

1.27

 vegetales

1.27

нее

1.27

先日

1.25

このような

1.24

Activations Density 0.035%