INDEX

Explanations

opinions, beliefs and disagreements.

variations of the word 'hold'

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

лтамалар

-0.57

zeba

-0.53

Rüyada

-0.52

 jsPsych

-0.52

findpost

-0.51

ьаж

-0.50

částí

-0.48

uttavia

-0.47

bondale

-0.47

sendRedirect

-0.46

POSITIVE LOGITS

 held

1.55

held

1.50

 hold

1.38

 holds

1.32

 holding

1.26

 Hold

1.26

Holding

1.25

 Holding

1.23

Hold

1.23

hold

1.23

Activations Density 1.692%