INDEX

Explanations

verbs after "to"

This neuron activates on “to” when it appears as part of instructional phrases (e.g. “don’t forget to,” “remember to,” “be sure to”), marking an imperative infinitive.

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

 only

-1.72

 will

-1.55

)」

-1.52

 both

-1.51

</sup>

-1.50

 traditionellen

-1.48

can

-1.47

 beendet

-1.47

 sicheren

-1.44

lysm

-1.43

POSITIVE LOGITS

 市场

1.46

ߣ

1.46

 网站

1.40

しも

1.37

药

1.37

 国家

1.35

绿

1.34

针

1.31

BIÉN

1.29

 veckor

1.28

Activations Density 0.002%