INDEX

Explanations

caused by

questions and follow-ups

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

נס

0.52

ент

0.51

тан

0.51

ಹಾರ

0.51

அந்த

0.50

yecto

0.50

atorul

0.49

После

0.49

прият

0.49

ОР

0.49

POSITIVE LOGITS

 examples

0.56

 drills

0.48

 hoodies

0.48

 sheeting

0.48

 samples

0.47

 benchmarks

0.47

 Examples

0.46

 Sheeran

0.46

 regulations

0.45

Activations Density 0.000%