INDEX

Explanations

future events and possibilities

The main thing this neuron does is detect phrasing that expresses anticipation of future events (e.g. “to come,” “may hold,” “in store”).

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

.",

-1.53

reason

-1.50

 بیشتری

-1.46

 yeni

-1.43

}$,

-1.34

 debat

-1.34

?”,

-1.32

enerys

-1.32

```

-1.32

\&

-1.32

POSITIVE LOGITS

 each

1.51

幸いです

1.45

any

1.45

in

1.43

 with

1.41

 Rather

1.40

 مختلف

1.35

舠

1.32

 Eventually

1.32

 this

1.30

Activations Density 0.022%