INDEX

Explanations

dates around May

The neuron fires on any standalone numeric token (including integers, decimals, IDs, years, etc.).

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

of

-1.44

fname

-1.40

</h4>

-1.39

『

-1.35

-1.34

warzys

-1.30

--

-1.30

-1.28

If

-1.25

entamiento

-1.24

POSITIVE LOGITS

 that

1.65

 those

1.64

</i>

1.53

厲害

1.44

épa

1.39

年は

1.34

čit

1.32

 some

1.32

 hardly

1.28

nými

1.26

Activations Density 0.005%