INDEX

Explanations

isolated from

scientific text describing samples isolated from specific locations or environments.

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

ícias

-0.93

 barca

-0.92

otropic

-0.84

ποι

-0.84

 kvar

-0.82

 speeds

-0.82

hilangan

-0.82

 instead

-0.82

 discoveries

-0.81

istos

-0.81

POSITIVE LOGITS

 samples

1.84

Samples

1.74

samples

1.48

 sampling

1.48

 Samples

1.48

 collected

1.45

 sampled

1.44

Sampling

1.35

 muestras

1.23

collected

1.23

Activations Density 0.292%