INDEX

Explanations

The neuron activates on document section titles or headlines (i.e. standalone, highly capitalized headings).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ませんので

0.71

<strong>

0.70

:"",

0.69

="",

0.67

that

0.66

='',

0.65

*,

0.64

%，

0.64

\",\"

0.63

input

0.63

POSITIVE LOGITS

</h2>

2.48

</h4>

2.28

</h3>

2.23

</h1>

1.99

</b>

1.83

</h5>

1.78

</strong>

1.67

‬

1.30

</td>

1.29

</h6>

1.23

Activations Density 0.506%