让AI辅助集体决策，既扩规模又保公平与话语权。
Accountable Human-AI Deliberation with LLMs: Scaling Collective Intelligence through Symbiotic Scaffolding
- 三层次框架：观察增多样性、带溯源的引导、人类最终拍板。
- 实测支持多视角表达，避免过度求同，保障参与感。
- 适合政策制定、公众议事等需公正透明的群体决策场景。
大语言模型（LLMs）可突破传统协商中轮次与协调能力的限制，实现大规模民主讨论。已有研究显示，由LLM生成的群体意见常优于人工引导输出，理论分析也表明其能缓解集体智能中的同步性约束。然而纯由LLM主导可能削弱多元性，过度追求共识，且当参与者无法质疑自身被如何代表时，会损害正当性。为此，我们提出一种人机共生框架，包含三层：观察与多样性增强、基于条款级溯源的引导、人类优先的确认机制。贡献包括分级覆盖、多样性与擦除度量（带显著性加权）；结合交叉编码器相似性与因果剔除诊断的溯源流水线；偏好条件下的权衡控制；公平敏感的争议处理流程；对抗鲁棒性测试；以及基于‘LLM作为裁判’局限性的消融设计评估协议。该方案提供了一个可验证的协商技术蓝图，在扩展集体智能的同时，确保个体自主性与合法性。
原文摘要 · Abstract (English)
Large language models (LLMs) can support democratic deliberation at scales previously constrained by turn-taking and facilitation bandwidth. Recent work shows that LLM-generated group statements are often preferred over human-mediated outputs, while theoretical analyses argue that LLMs relax the simultaneity constraints limiting collective intelligence. Yet pure LLM mediation risks collapsing pluralism, over-optimizing for agreement, and undermining legitimacy when participants cannot contest how they are represented. We propose a symbiotic human-AI framework organized into three layers: observation and diversity amplification, facilitation with clause-level provenance, and human primacy for ratification. Our contributions include graded coverage, diversity, and erasure metrics with salience-aware weighting; a provenance pipeline combining cross-encoder similarity with causal knockout diagnostics; preference-conditioned trade-off control; equity-aware contestability workflows; adversarial robustness tests; and an evaluation protocol with ablation designs informed by evidence of LLM-as-judge limitations. The result is a testable blueprint for deliberation technology that scales collective intelligence while preserving agency and legitimacy.
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。