CDCK/MoM приводит посты от User1, User2 и т.д. — проблема с качеством LLM?

Falco · 10.Август.2026 16:49:01

Спасибо за сообщение. Эта PR решает данную проблему:

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.

Теперь сводка по этой теме выглядит лучше.

Тема		Ответов	Просм.
Moderation tool for formatting code with AI Feature ai	10	425	03.04.2025
Code fences don't seem to deal with C comments Support	26	4841	06.10.2015
HTML comments are also summarized by AI Feature ai , ai-summarize	5	326	01.07.2026
Pasted content gains erroneous line breaks Bug needs-repro	5	263	19.08.2024
Catch and educate users posting code without properly formatting it Feature	28	5597	11.04.2019

CDCK/MoM приводит посты от User1, User2 и т.д. — проблема с качеством LLM?

Связанные темы