INDEX

Explanations

asking questions

posing questions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 disagreements

0.57

 disagree

0.52

 debates

0.51

 disagreement

0.51

 كلام

0.51

 disputes

0.50

 disagreed

0.50

 Disputes

0.48

 controversies

0.48

 Dispute

0.48

POSITIVE LOGITS

 question

0.89

 pytanie

0.83

 pitanje

0.79

 pergunta

0.76

 otáz

0.73

 pregunta

0.73

question

0.72

 wondered

0.71

 spør

0.71

 вопрос

0.70

Activations Density 0.076%