INDEX

Explanations

computer system

predicting predictions and assessments

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

 الصورة

0.52

 Tensor

0.52

 Border

0.47

 Illusion

0.47

 सक्रिय

0.46

 Modulation

0.45

 Caul

0.44

 Telescope

0.44

 Pathol

0.44

 Sousa

0.44

POSITIVE LOGITS

ни

0.57

ме

0.56

не

0.52

વ

0.51

Д

0.50

ચ

0.49

APPLE

0.48

agree

0.48

Mathematical

0.48

 typographical

0.48

Activations Density 0.001%