INDEX

Explanations

auxiliary verbs and modals

The neuron strongly activates on auxiliary verbs (e.g. “re,” “is,” “have,” “will,” “need”) indicating it detects helper verbs rather than content words.

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

of

-1.45

 інших

-1.38

إنه

-1.35

These

-1.27

高温

-1.25

mathrm

-1.21

Artik

-1.21



-1.20

reifen

-1.20

其他

-1.20

POSITIVE LOGITS

 souvent

1.38

potentially

1.27

 Reſ

1.25

 bricola

1.23

 IBOutlet

1.23

蹿

1.23

became

1.22

耵

1.20

了一个

1.20

𓆏

1.19

Activations Density 0.082%