INDEX

Explanations

questions starting with why did

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

0.85

 prof

0.76

0.74

 жало

0.73

vol

0.73

st

0.70

hr

0.69

fea

0.69

!;

0.67

fel

0.66

POSITIVE LOGITS

Calculation

1.53

Calculate

1.52

Calculating

1.50

Alright

1.50

Numerator

1.47

Exponent

1.44

Exponential

1.44

Relative

1.42

Ignore

1.42

Computing

1.41

Activations Density 0.001%