INDEX

Explanations

won't/wouldn't + refuse/prevent verbs

The neuron is highly sensitive to auxiliary and modal verbs (like “would,” “will,” “won’t”) indicating possibility, intention, or refusal.

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

 graciosos

-1.77

ݮ

-1.56

 graciosas

-1.56

 cristianos

-1.54

 religiosos

-1.50

JAJA

-1.47

ciuto

-1.45

 votre

-1.42

 creativos

-1.42

งิน

-1.41

POSITIVE LOGITS

is

1.49

1.37

and

1.32

1.29

1.24

</b>

1.24

 currently

1.23

 following

1.23

</strong>

1.22

er

1.22

Activations Density 0.035%