INDEX

Explanations

references to monitoring or measurement activities

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 imageNamed

-0.59

baix

-0.56

 monitor

-0.54

 lari

-0.54

 fres

-0.52

ẢN

-0.50

tvguidetime

-0.50

hebt

-0.49

 Monitor

-0.48

tagli

-0.47

POSITIVE LOGITS

 monitored

1.97

itored

1.11

 supervised

0.93

UnusedPrivate

0.90

 tracked

0.78

 kaynağından

0.72

kloped

0.71

parsedMessage

0.70

 oversaw

0.69

Datuak

0.66

Activations Density 0.003%