INDEX

Explanations

am i, assume with, have we

The neuron fires on hedging or disclaimer phrases—particularly “in no way am I making a statement (about …)”–style disclaimers.

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

 todėl

-1.58

 поэтому

-1.44

².

-1.36

了不少

-1.34

 때문

-1.30

 publicaciones

-1.26

 vitales

-1.24

 anunció

-1.22

;</

-1.21

了你

-1.21

POSITIVE LOGITS

 from

1.85

 including

1.75

one

1.73

the

1.69

 regarding

1.47

at

1.43

 that

1.37

 without

1.37

 aspects

1.30

 other

1.27

Activations Density 0.199%