INDEX

Explanations

ASCII table borders

The neuron fires on numeric literals (including numbers in format strings and numeric constants in code).

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

it

-1.62

or

-1.52

how

-1.49

 that

-1.48

of

-1.45

 While

-1.37

 what

-1.36

 only

-1.27

 without

-1.27

 this

-1.27

POSITIVE LOGITS

などを

1.23

そして

1.22

があり

1.12

を

1.06

 melding

1.06

 ได้

1.05

 promulg

1.05

撵

1.04

 McDonalds

1.04

 leurs

1.04

Activations Density 0.003%