INDEX

Explanations

code comments and headers

The neuron strongly activates on long runs of asterisks (i.e. banner‐style comment separators).

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

 společnost

-1.88

⬧

-1.81

他们是

-1.80

you

-1.76

 seguras

-1.73

 motivadoras

-1.73

襖

-1.72

मति

-1.63

jedi

-1.63

 surgió

-1.62

POSITIVE LOGITS

</h3>

2.63

</i>

2.33

.";

2.09

на

1.95

1.88

only

1.77

ностей

1.70

most

1.70

م

1.66

";

1.63

Activations Density 0.010%