INDEX

Explanations

verbs following 'to'

The neuron strongly activates on the token “to,” i.e. the infinitive‐marker “to” in text.

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

蹰

-1.44

 שלנו

-1.42

 フォト

-1.38

).

-1.38

櫪

-1.35

 tradisi

-1.31

 matchs

-1.30

 appris

-1.29

 feita

-1.27

↱

-1.27

POSITIVE LOGITS

 this

1.70

 they

1.63

begin

1.53

to

1.38

と

1.32

to

1.30

 approach

1.30

十分钟

1.27

acaktır

1.27

gelopen

1.26

Activations Density 0.051%