INDEX

Explanations

rise/raise and related words, often followed by "up" or indicating increases/concerns

The neuron fires on occurrences of the verb “raise” (and its inflected forms) wherever something is lifted or increased.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

郯

0.66

 Pohl

0.59

λω

0.59

Prints

0.59

 FIELDS

0.58

 Jump

0.56

 Watanabe

0.56

माना

0.55

hong

0.55

 Places

0.54

POSITIVE LOGITS

 raising

1.48

 raise

1.46

 raises

1.38

Raise

1.38

raising

1.38

raised

1.38

raise

1.36

 raised

1.33

 подня

1.29

Raising

1.27

Activations Density 0.212%