progressive-citation-grounded-dialogue

【开发与工程】Eliminate hallucination via four-stage progressive training: multilingual adaptation → English dialogue SFT → bilingual SFT → GRPO alignment. Achieve 0.0% hallucination rate for encoder-decoder models using structured citation markers and knowledge-source attribution, with automatic transfer of citation format across languages.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准