INDEX

Explanations

working with

The neuron detects participial or prepositional method‐oriented phrases (e.g. “Using …,” “working in …,” “on a lightly floured surface, …”) that introduce procedural or methodological steps.

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

㌘

-1.42

㈯

-1.37

 nimmer

-1.34

 compró

-1.33

嘍

-1.28

 redução

-1.27

 richiede

-1.26

uhh

-1.25

 transmissão

-1.25

 possibilità

-1.23

POSITIVE LOGITS

of

1.57

by

1.38

 through

1.35

 without

1.30

 which

1.29

or

1.27

 misguided

1.27

pplung

1.27

 Going

1.23

・)

1.20

Activations Density 0.093%