INDEX

Explanations

can be + adjective/past participle

The neuron strongly activates on occurrences of the auxiliary verb “be” (especially in “can be” constructions) and similar high-frequency modal/helper verbs.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

-1.50

-1.43

ish

-1.38

jdź

-1.36

んだから

-1.34

hdashline

-1.32

 значит

-1.31

 about

-1.29

va

-1.28

，「

-1.27

POSITIVE LOGITS

 However

1.66

we

1.59

 данный

1.51

 sogar

1.41

 Этот

1.40

 նաև

1.39

you

1.37

 Practically

1.32

You

1.30

 erzä

1.30

Activations Density 0.057%