INDEX

Explanations

I was feeling

The neuron is detecting first-person self-references (e.g. “I,” “my,” verbs of personal actions).

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

you

-1.66

for

-1.43

 from

-1.42

 your

-1.29

You

-1.26

就是这样

-1.23

 each

-1.20

 всички

-1.19

All

-1.19

Two

-1.13

POSITIVE LOGITS

 “...

1.43

 “…

1.34

 "...

1.27

 probablement

1.24

🙀

1.22

楫

1.20

‘

1.18

“

1.18

fois

1.15

város

1.15

Activations Density 0.042%