用小模型组队互评，让边缘端问答更靠谱。
EdgeJury: Cross-Reviewed Small-Model Ensembles for Truthful Question Answering on Serverless Edge Inference
- 四阶段协作：角色生成、匿名互评、主席整合、一致性标注
- 在TruthfulQA上达76.2%准确率，比单模型高21.4%
- 适合资源受限的边缘部署，无需大模型或外部检索
幻觉问题严重制约可靠问答，尤其在资源受限的边缘场景中，前沿大模型或检索系统难以落地。本文提出EdgeJury，一种轻量级小模型集成框架，仅使用3B-8B参数的指令微调语言模型，适配无服务器边缘推理。该框架包含四个阶段：(1) 并行角色专业化生成，(2) 带结构化批评与排名的匿名交叉评审，(3) 主席合成整合强内容并修正问题，(4) 基于多模型一致性的声明级一致性标注。在TruthfulQA（MC1）上，EdgeJury达到76.2%准确率（95%置信区间：72.8-79.6%），相较单个8B基线（62.8%）提升21.4%，优于自一致性与多数投票等标准方法，并公开计算成本（总令牌数与平台开销）。在200题对抗性EdgeCases测试集上，相对提升达48.2%（95%置信区间：44.0-52.4%）。人工分析100个错误答案显示，事实幻觉错误减少约55%。部署于Cloudflare Workers AI，端到端中位延迟为8.4秒，证明协同小模型集成可在无需外部检索或专有大模型接口下，显著提升误解密集型问答任务的可信度。
原文摘要 · Abstract (English)
Hallucinations hinder reliable question answering, especially in resource-constrained deployments where frontier-scale models or retrieval pipelines may be impractical. We present EdgeJury, a lightweight ensemble framework that improves truthfulness and robustness using only small instruction-tuned language models (3B-8B) suitable for serverless edge inference. EdgeJury orchestrates four stages: (1) parallel role-specialized generation, (2) anonymized cross-review with structured critiques and rankings, (3) chairman synthesis that integrates the strongest content while addressing flagged issues, and (4) claim-level consistency labeling based on inter-model agreement. On TruthfulQA (MC1), EdgeJury achieves 76.2% accuracy (95% CI: 72.8-79.6%), a +21.4% relative improvement over a single 8B baseline (62.8%), and outperforms standard baselines including self-consistency and majority voting under transparent compute accounting (total tokens and platform cost reported). On a 200-question adversarial EdgeCases set, EdgeJury yields +48.2% relative gains (95% CI: 44.0-52.4%). Manual analysis on 100 incorrect answers shows an approximately 55% reduction in factual hallucination errors versus the single-model baseline. Deployed on Cloudflare Workers AI, EdgeJury achieves 8.4 s median end-to-end latency, demonstrating that coordinated small-model ensembles can improve truthfulness on misconception-heavy QA benchmarks without external retrieval or proprietary large-model APIs.
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。