INDEX

Explanations

+- and *- symbols

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

+-

-0.92

+-

-0.78

 Hote

-0.70

°-

-0.68

 expectancy

-0.59

cer

-0.55

 hote

-0.55

-0.52

KeyValuePair

-0.52

 consta

-0.51

POSITIVE LOGITS

fromnode

0.73

Биография

0.62

parsedMessage

0.59

 AssemblyTitle

0.58

 AssemblyCulture

0.56

SourceChecksum

0.54

posedge

0.54

 HttpNotFound

0.54



0.51

Activations Density 0.069%