Left examples have the keyword "literalgeometry" on the OEBP.

As a default, it is assumed on the OEBP that submitted examples are imperfect drawings. A problem about circles may only feature what are approximately circles in its examples. Thus, most solutions to Bongard Problems have the implicit caveat "...if shapes in drawings are parsed as what they most intuitively approximate." In literalgeometry problems, this caveat is dropped; the solution is based on the literal geometry of the drawings used as examples, including all subtle imperfections.

Bitmaps can only hold so much information, so even in literalgeometry Problems it is not typically expected that the solver will take into account pixelation of images. (Problems in which pixelation is relevant are usually assigned the world "bmp" on the OEBP.) In literalgeometry Problems having to do with smooth curves and lines using pixelated images such as .png, .gif, .bmp, and .jpg files, it is only necessary that the pixelated images offer the best possible approximation of the underlying shape given their resolutions. On the other hand, if the OEBP would indeed prefer users only upload pixel-perfect examples, a Problem can be tagged with the keyword "pixelperfect" (left-BP947).

Regardless of whether in literalgeometry Problems or not, .svg scalable vector graphics files are always assumed to be parsed as their literal geometry. Do not upload .svg files with imperfections, unless the BP solution is explicitly about those imperfections.

Just because a Problem involves small details does not make it "literalgeometry." It is possible for a Problem to involve small imperfections, but still be about the intuitive parsings of drawings rather than the literal geometry in the drawings, e.g. BP148 or BP119.

See BP508 for more discussion of this.

Left examples have the keyword "literalgeometry" on the OEBP.

As a default, it is assumed on the OEBP that submitted examples are imperfect drawings. A problem about circles may only feature what are approximately circles in its examples. Thus, most solutions to Bongard Problems have the implicit caveat "...if shapes in drawings are parsed as what they most intuitively approximate." In literalgeometry problems, this caveat is dropped; the solution is based on the literal geometry of the drawings used as examples, including all subtle imperfections.

Bitmaps can only hold so much information, so even in literalgeometry Problems it is not typically expected that the solver will take into account pixelation of images. (Problems in which pixelation is relevant are usually assigned the world "bmp" on the OEBP.) In literalgeometry Problems having to do with smooth curves and lines using pixelated images such as .png, .gif, .bmp, and .jpg files, it is only necessary that the pixelated images offer the best possible approximation of the underlying shape given their resolutions. On the other hand, if the OEBP would indeed prefer users only upload pixel-perfect examples, a Problem can be tagged with the keyword "pixelperfect" (left-BP947)

Regardless of whether in literalgeometry Problems or not, .svg scalable vector graphics files are always assumed to be parsed as their literal geometry. Do not upload .svg files with imperfections, unless the BP solution is explicitly about those imperfections.

Just because a Problem involves small details does not make it "literalgeometry." It is possible for a Problem to involve small imperfections, but still be about the intuitive parsings of drawings rather than the literal geometry in the drawings, e.g. BP148 or BP119.

See BP508 for more discussion of this.