INDEX

Explanations

influence and inspiration

The main thing this neuron does is detect mentions of influences or inspiration (e.g., “influences,” “inspired,” “inspiration”).

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

ּוֹ

-1.34

Переваги

-1.30

ַּ

-1.23

ִּ

-1.18

倖

-1.17

当時は

-1.11

lova

-1.09

 Louvain

-1.08

ございません

-1.08

}}}$

-1.05

POSITIVE LOGITS

 from

1.95

 influences

1.87

 inspiration

1.66

 inspirations

1.61

Inspiration

1.49

 beeinf

1.44

inspiration

1.41

 Inspiration

1.39

 influence

1.38

 sources

1.33

Activations Density 0.032%