INDEX

Explanations

understand and know how

This neuron detects occurrences of the word “understand” (especially in the common phrase “we understand that…” or “we understand the…”) in marketing or corporate‐speak contexts.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

two

-1.79

 without

-1.49

外国人

-1.41

 χωρίς

-1.40

日発売

-1.39

 zonder

-1.38

 概念

-1.34

Without

-1.34

 Without

-1.33

ácidos

-1.33

POSITIVE LOGITS

you

2.02

 your

1.75

Ú

1.45

在

1.45

ୌ

1.41

 which

1.37

 DbSet

1.34

 onAnimation

1.33

柃

1.31

篭

1.30

Activations Density 0.017%