INDEX

Explanations

expressions of identifying and highlighting mistakes or errors

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

No Comments

Negative Logits

ElementException

-0.64

èze

-0.57

MathML

-0.56

ErrorListener

-0.54

ichthys

-0.53

Географи

-0.52

++]=

-0.52

Външни

-0.52

providedIn

-0.52

 equivalence

-0.50

POSITIVE LOGITS

 mistake

2.09

 Mistake

1.45

Mistake

1.13

 mistakes

0.98

 Mistakes

0.82

تقاوى

0.72

 hObject

0.66

:✨

0.64

 sacrificing

0.62

 mistaken

0.61

Activations Density 0.007%