INDEX

Explanations

table elements

This neuron detects HTML table cell tags (i.e. `<td>` or `</td>` elements).

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

or

-1.82

is

-1.78

may

-1.70

 even

-1.69

 also

-1.67

of

-1.59

-1.52

 actually

-1.48

by

-1.45

 could

-1.45

POSITIVE LOGITS

 différents

1.48

obviously

1.38

enjoy

1.38

unfortunately

1.34

strongly

1.34

 разных

1.32

forgot

1.31

螃

1.31

unexpected

1.30

midler

1.30

Activations Density 0.011%