INDEX

Explanations

using tools or software

The neuron is keyed to the instructional verb “using” (and closely related tool-use contexts), marking phrases that describe how something is done.

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

 asian

-1.60

 lorenzo

-1.46

 greek

-1.42

而

-1.41

 frankfurt

-1.41

而且

-1.40

 giorgio

-1.34

但是

-1.29

 другое

-1.27

 milf

-1.27

POSITIVE LOGITS



1.70

 maravillosa

1.55

ܜ

1.52



1.50

 stronie

1.48

 сумка

1.46

tifies

1.45

熺

1.38

 lindas

1.37

nosed

1.35

Activations Density 0.000%