INDEX

Explanations

numbers followed by a hyphen

The neuron is primarily responding to tokens that consist of standalone numbers or short uppercase/alphanumeric codes (e.g. digits and acronyms).

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

-1.66

but

-1.48

 only

-1.36

any

-1.33

 even

-1.31

There

-1.31

 after

-1.30

 there

-1.29

 späteren

-1.27

 with

-1.26

POSITIVE LOGITS

 ..............

1.56

 ............

1.38

 horloge

1.36

 ...........

1.35

 .........

1.35

 ........

1.34

 xvii

1.33

 ..........

1.31

 applau

1.30

 ovale

1.29

Activations Density 0.465%