INDEX

Explanations

The neuron responds to numeric tokens (digits), especially those appearing in dates or version numbers.

URLs and post titles

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

for

-1.75

but

-1.75

if

-1.44

If

-1.44

With

-1.43

 While

-1.40

can

-1.40

 However

-1.37

or

-1.34

While

-1.31

POSITIVE LOGITS

 ARISING

1.43

︐

1.38

缡

1.34

螃

1.34

܆

1.33

还不错

1.31

憷

1.27

 AstraZeneca

1.27

raires

1.27

ensitive

1.26

Activations Density 0.008%