INDEX

Explanations

code syntax and keywords

The neuron is highly sensitive to any token containing digits (numeric literals or numbers) in code.

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

 berlina

-0.82

 soiled

-0.80

ári

-0.75

 pitfalls

-0.75

 lupa

-0.74

المراجع

-0.73

 kotor

-0.71

MUR

-0.71

 kanan

-0.70

Guyana

-0.70

POSITIVE LOGITS

modelBuilder

0.87

 ObjectParameter

0.84

unnitel

0.79

 nhà

0.78

 Kinect

0.77

cellation

0.77

 jedoch

0.76

hob

0.76

cee

0.73

創業

0.73

Activations Density 0.003%