INDEX

Explanations

let followed by mathematical expression

This neuron fires on introductory assumption or definition words (e.g. “Let,” “Suppose”) that begin a mathematical statement.

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

ilistic

-1.34

ጐ

-1.33

了不少

-1.32

etration

-1.28

bourhood

-1.25

 keinerlei

-1.22

ungguh

-1.22

Aplica

-1.20

他们

-1.18

他們

-1.18

POSITIVE LOGITS

the

1.86

 ternyata

1.35

примеча

1.33

 there

1.26

we

1.21

1.20

 also

1.18

coglie

1.17

in

1.14

キリスト

1.08

Activations Density 0.043%