CDCK/MoM zitiert Beiträge als User1, User2 usw. – LLM-Qualitätsproblem?

Falco · 10. August 2026 um 16:49

Danke für den Bericht, dieser PR behebt dieses Problem:

github.com/discourse/discourse

DEV: Use a realistic few-shot example in the summarizer agent (#42445)

main ← dev-summarizer-realistic-example

merged 06:41PM - 07 Aug 26 UTC

xfalcox

+14 -3

Previously, the summarizer agent's few-shot example used a two-post toy thread w…hose format didn't match the prompts the summarization strategies actually build, and the system prompt described a post format (`N) username message`) that no strategy emits. This change replaces the example with one mirroring a real captured request (discussion title line, posts inside `<input></input>` tags in the real `(N username said: ...` format) paired with a multi-paragraph cited summary demonstrating `[username]({resource_url}/N)` citations, and fixes the post format description to match what the topic, chat, and gist strategies actually send.

Die Zusammenfassung zu diesem Thema sieht jetzt besser aus.

Thema		Antworten	Aufrufe
Moderation tool for formatting code with AI Feature ai	10	330	3. April 2025
Code fences don't seem to deal with C comments Support	26	4716	6. Oktober 2015
HTML comments are also summarized by AI Feature ai , ai-summarize	5	220	1. Juli 2026
Pasted content gains erroneous line breaks Bug needs-repro	5	201	19. August 2024
HTML/RTF pasting Feature	74	13776	9. März 2020

CDCK/MoM zitiert Beiträge als User1, User2 usw. – LLM-Qualitätsproblem?

Verwandte Themen