INDEX

Explanations

This neuron detects sentence-ending boundaries (periods marking the ends of sentences).

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

</em>

1.06

</i>

0.88

}\

0.80

})

0.75

}=

0.75

}=\

0.74

，

0.74

}}$

0.73

}-

0.68

))

0.67

POSITIVE LOGITS

<h3>

4.31

</h3>

2.61

"];

2.41

###

2.25

subsubsection

2.23

'];

2.17

###

2.11

<h4>

2.03

↵↵↵↵↵

1.80

</th>

1.80

Activations Density 0.185%