INDEX

Explanations

dates

The main thing this neuron does is detect instances of the word “dates,” typically when labeling or listing event or schedule dates.

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

</td>

-2.45

ים

-2.23

</i>

-2.16

を作る

-2.11

rebbero

-2.08

也

-2.06

 archivos

-2.05

にも

-2.03

 με

-2.00

 stench

-1.98

POSITIVE LOGITS

of

2.59

an

2.44

ly

2.28

 dates

2.16

時間も

1.99

1.96

</strong>

1.95

 cites

1.90

 creando

1.85

ၞ

1.84

Activations Density 0.004%