INDEX

Explanations

company endings

The neuron fires most strongly on words in instructional or explanatory passages—i.e. content giving advice, directions, or describing how to do something.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

↵↵↵↵

0.86

↵↵↵↵↵

0.84

<b>

0.82

0.80

<strong>

0.77

<unused61>

0.75

↵↵↵

0.67

);//

0.65

↵↵↵↵↵↵

0.64

";//

0.64

POSITIVE LOGITS

 tional

1.02

 ción

0.96

 tions

0.94

 tion

0.92

 tive

0.82

 ments

0.79

 അദ്ദേ

0.79

ity

0.74

ous

0.72

 }}^{\

0.70

Activations Density 0.451%