INDEX

Explanations

meet up with

The neuron detects mentions of gatherings or get-togethers, i.e. “meet,” “meet up,” “meet-ups,” and “meeting.”

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

to

-2.23

!”

-2.22

how

-2.22

”,

-2.16

.’

-2.03

—

-2.03

!).

-1.89

-1.78

but

-1.77



-1.76

POSITIVE LOGITS

蕻

2.06

寕

2.03

 privés

2.03

 (...

1.96

?...

1.96

 rendus

1.96

樒

1.95

 sustains

1.88

苼

1.86

 acknowledges

1.80

Activations Density 0.002%