INDEX

Explanations

for purposes

The neuron activates on verbs that take the preposition “for” (e.g. apply for, ask for, pay for, bid for, make fun of).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ی

1.95

ли

1.52

IA

1.50

ా

1.50

์

1.44

دت

1.39

ో

1.39

罗

1.38

ه

1.38

UR

1.37

POSITIVE LOGITS

giveness

2.38

asmuch

1.86

ations

1.77

 phép

1.72

ac

1.71

 purposes

1.66

ри

1.65

 Damages

1.48

 awhile

1.44

aying

1.44

Activations Density 1.197%