INDEX

Explanations

myths misconceptions stereotypes

The neuron activates on words that denote misconceptions, myths, stereotypes, or the act of debunking false beliefs.

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

 from

-1.01

楽曲

-0.97

Topology

-0.92

entuh

-0.89

to

-0.88

 nostal

-0.88

for

-0.88

 foreshadow

-0.87

 without

-0.85

onPressed

-0.85

POSITIVE LOGITS

 about

2.28

 myths

1.69

 misconceptions

1.53

 stereotypes

1.35

myth

1.32

 rằng

1.29

 belief

1.27

 Myths

1.25

關於

1.25

 myth

1.25

Activations Density 0.040%