INDEX

Explanations

accessible from

The neuron fires on phrasing that something “can be accessed/reached/seen from” a given place or interface.

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

つまり

-1.41

⚋

-1.30

 respectivas

-1.28

↵↵

-1.27

</strong>

-1.26

}^{\

-1.25

↵↵↵

-1.24

</h1>

-1.20

 gratuito

-1.20

幹線

-1.20

POSITIVE LOGITS

its

1.88

 your

1.78

1.59

the

1.55

我一

1.48

去吧

1.39

調べて

1.34

鉀

1.30

﹐

1.30

 régiment

1.29

Activations Density 0.064%