INDEX

Explanations

just over or under

The neuron is tuned to detect approximate time‐duration phrases, e.g. “just over a year.”

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

-2.48

The

-1.78

and

-1.69

to

-1.69

 What

-1.67

Ꮈ

-1.63

 paramount

-1.58

 comprando

-1.57

仮

-1.48

POSITIVE LOGITS

1.82

</tfoot>

1.81

 tragé

1.80

 bús

1.77

 palha

1.70

ltä

1.70

꜆

1.69

zieć

1.62

 péro

1.61

contentValues

1.61

Activations Density 0.004%