INDEX

Explanations

words related to quality, manufacturing processes, and health, or words used in a modal phrase

suffer

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

спла

-0.55

abestanden

-0.51

 برانيه

-0.50

 africains

-0.48

brug

-0.46

 verteilt

-0.46

 bezeichneter

-0.45

 initComponents

-0.44

Sziasztok

-0.43

 volna

-0.43

POSITIVE LOGITS

 suffer

1.30

 suffers

1.21

 Suffer

1.10

 suffering

1.05

 suffered

0.98

 Suffering

0.97

suffering

0.96

 Roskov

0.88

 sufferings

0.78

 sufferer

0.71

Activations Density 1.775%