INDEX

Explanations

numbers and measurements

This neuron specializes in spotting standalone numeric tokens that denote technical specifications (e.g. capacities, model numbers, dimensions).

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

つづく

-1.73

色々と

-1.66

様な

-1.52

様に

-1.50

うわ

-1.38

あぁ

-1.38

久し

-1.35

 évit

-1.33

 拡大

-1.30

いきなり

-1.30

POSITIVE LOGITS

of

2.16

in

1.52

but

1.48

utenants

1.38

 could

1.37

 because

1.36

 that

1.35

を

1.35

 would

1.35

now

1.34

Activations Density 0.035%