INDEX

Explanations

installed on

predictable performance, dependencies

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

 regulación

0.55

 Regulation

0.51

 regulation

0.49

 विनियमन

0.48

 দ্বি

0.46

<0xAB>

0.46

 qualifications

0.45

MÉ

0.45

 EXAMINATION

0.44

À

0.44

POSITIVE LOGITS

0.57

care

0.53

extensions

0.51

 istem

0.50

ACTION

0.49

illegal

0.49

leben

0.48

consistent

0.48

done

0.48

action

0.47

Activations Density 0.001%