INDEX

Explanations

let me know

This neuron detects polite request phrases—especially “please let me know” and similar “help me”/“let me know” solicitations.

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

to

-2.22

ReusableCell

-1.80

tività

-1.73

rogates

-1.58

︲

-1.57

LayoutConstraint

-1.55

 again

-1.55

}{*}{}

-1.52

佧

-1.52

tainty

-1.51

POSITIVE LOGITS

↵↵

1.36

任何

1.34

 ساعد

1.33

</h1>

1.32

lež

1.30

1.28

 его

1.27

 resistente

1.23

無法

1.20

ле

1.18

Activations Density 0.035%