INDEX

Explanations

instances of "the" followed by a word indicating a generally positive result

the

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

Negative Logits

#+#

-0.77

DropTable

-0.60

expandindo

-0.60

dropIfExists

-0.55

 AssemblyProduct

-0.51

sidemargin

-0.51

 getItemId

-0.50

 Hinduism

-0.49

dolu

-0.48

édrale

-0.48

POSITIVE LOGITS

']")

0.61

kloped

0.52

 rest

0.51

(||

0.50

devamını

0.49

Parcelize

0.48

 toughest

0.47

⃣

0.47

acağı

0.47

ExtendWith

0.46

Activations Density 0.567%