News for hackers. Stuff that matters.

A comment on GLM-5.2 is the new leading open weights model on Artificial Analysis. Parent comment

Re:GLM-5.2 is the new leading open weights model on Artifici... by saltsucker on Wednesday June 17, 2026 @09:48AM (#48570554)
Comments later in thread say markdown works just as fine and that it’s more important to organize your plan into sections.

Also just think about it, why would a model trained on the world’s corpus of text (that isnt formatted in xml) perform better with XML? It would be a better study if that post tested markdown, org, xml, json, etc. 10 times to see if their is a difference

[ Parent | on HN ]
Re:GLM-5.2 is the new leading open weights model on Artifici... by swingboy on Wednesday June 17, 2026 @11:53AM (#48572213)
Anthropic’s best practices still include the use of XML: https://platform.claude.com/docs/en/build-with-claude/prompt...
[ Parent | on HN ]
Re:GLM-5.2 is the new leading open weights model on Artifici... by adastra22 on Wednesday June 17, 2026 @11:08AM (#48571578)
A year or so ago XML worked more reliably for long-lived prompt instructions. Now it is cargo culting.
[ Parent | on HN ]
Re:GLM-5.2 is the new leading open weights model on Artifici... by orbital-decay on Wednesday June 17, 2026 @06:15PM (#48577700)
XML consistently performed better than markdown and JSON in all evals I've ever seen on any model, except for a couple very specific ones.
[ Parent | on HN ]