NightFeats通过三阶段协作机制提升文本生成的可解释性与可信度。
NightFeats @ MMU-RAGent NeurIPS 2025: A Context-Optimized Multi-Agent RAG System for the Text-to-Text Track
- 分三阶段：检索、筛选、合成，每阶段有明确接口与中间表示。
- 在人类评分和模型判别中优于Claude-SonnetV2等商用系统。
- 适合关注可解释性与事实准确性的AI系统研发者。
我们提出NightFeats，一个为NeurIPS 2025 MMU-RAGent竞赛设计的结构化多智能体检索增强生成系统，在文本到文本赛道获得最佳动态评估奖。不同于追求基准分数最大化，该工作提出一种原则性流程，将知识合成分解为检索、筛选与组合三个协同阶段，每个阶段由显式中间表示和交接契约控制。受Agentic Context Engineering（ACE）启发，系统引入时序语义重排序、受限矛盾消解与引用保持型组合作为核心架构组件。竞赛结果表明，NightFeats在LLM-as-a-Judge与人工李克特评分中超越包括Claude-SonnetV2和Nova-Pro在内的多个商业基线，证实架构透明性与可验证的事实锚定比仅优化自动相似度指标更契合人类偏好。
原文摘要 · Abstract (English)
We present NightFeats, a structured multi-agent retrieval-augmented generation (RAG) system submitted to the MMU-RAGent competition at NeurIPS 2025, where it was awarded Best Dynamic Evaluation in the text-to-text track. Rather than targeting benchmark maximization, this work proposes a principled pipeline that decomposes knowledge synthesis into three coordinated phases: retrieval, curation, and composition, each governed by explicit intermediate representations and handoff contracts. Inspired by Agentic Context Engineering (ACE), the system introduces temporal-semantic reranking, bounded contradiction reconciliation, and citation-preserving composition as core architectural primitives. Competition results show that NightFeats surpasses proprietary baselines including Claude-SonnetV2 and Nova-Pro on LLM-as-a-Judge and Human Likert evaluations, confirming that architectural transparency and verifiable evidence grounding are better aligned with human preferences than systems optimizing narrowly for automatic similarity metrics.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。