INDEX

Explanations

what about

This neuron fires on the “What… about…” or “Things You Need to Know About…” style phrasing in article headlines and questions.

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

鸶

-1.60



-1.52

ziek

-1.49

itettura

-1.45

皛

-1.43

に行ってきました

-1.41

泺

-1.41

茀

-1.38

uttgart

-1.34

'))

-1.34

POSITIVE LOGITS

from

2.03

1.73

the

1.59

isesta

1.50

how

1.48

him

1.47

都能

1.42

 FROM

1.41

 triliun

1.39

Activations Density 0.034%