INDEX

Explanations

what to use/spend/risk

The neuron detects the main action verb in questions (e.g. “help,” “use,” “make,” “be,” “look for”).

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

のだろうか

-1.78

硨

-1.77

𝒂

-1.76

蹰

-1.75

呖

-1.73

 イラストレーション

-1.73

龊

-1.73

 seulement

-1.71

裥

-1.67

 mauva

-1.65

POSITIVE LOGITS

for

2.13

 makes

2.05

 which

2.05

 When

1.98

 after

1.95

 showed

1.84

 went

1.80

 nine

1.80

 most

1.80

 came

1.77

Activations Density 0.049%