INDEX

Explanations

words that are prepositions, conjunctions, some forms of "to be," or that follow prepositions, suggesting that it finds grammatical connections between words.

terms related to respect and regard

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

Pautan

-0.58

 cuales

-0.50

 voulu

-0.48

kdy

-0.46

 środ

-0.46

PreferredItem

-0.45

gdx

-0.45

 Whether

-0.44

 Stunde

-0.43

<u>

-0.43

POSITIVE LOGITS

 respect

1.11

 impunity

1.06

 regard

1.05

 regards

1.02

regard

0.91

 ease

0.91

respect

0.87

inthe

0.84

]--;

0.82

outa

0.81

Activations Density 16.235%