发表机构
Zhejiang University; Inclusion AI, Ant Group; Peking University; Qwen Team, Alibaba Group（浙江大学; 蚂蚁集团Inclusion AI; 北京大学; 阿里巴巴集团通义千问团队）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文针对LLM的HSS数据稀缺问题，提出以学科为中心的HSS-Synth流水线，生成23.7万指令微调样本，使Qwen3-8B-Base达SOTA并提升相关能力。
AI 中文摘要
高质量、多样化的数据对大语言模型（LLM）至关重要，但仍稀缺且成本高昂。数据合成是可行的替代方案，在封闭任务中表现出色，但人文社科（HSS）领域被忽视，其开放性特质使合成工作颇具挑战。不同于以往以能力为核心、零散的尝试，本文采用以学科为中心的范式，定义了首个涵盖14个主流领域的HSS领域体系，并推出首个面向HSS的数据合成流水线HSS-Synth。HSS-Synth包含三部分：（1）通过多步骤过滤和由评判者评估的文本优化，从网络语料库构建种子文档；（2）指定“需求+人设”，将种子文档回译为多样化但忠实的指令，并进行严格的问答对齐检查；（3）通过教师强制回答突破LLM响应限制，在响应生成过程中输入种子文档以锚定语义、减少幻觉、保留语气与完整性。HSS-Synth生成了23.7万高质量、多样化的指令微调样本，在16个基准测试中优于14个领先基线。微调后的Qwen3-8B-Base达到新的SOTA，接近官方Qwen3-8B，在无性能波动的情况下同时提升了人类偏好和知识能力。大量实验验证了HSS-Synth的鲁棒性与可迁移性，其代码公开于此https URL。
英文摘要
High-quality, diverse data are vital for large language models (LLMs) but remain scarce and costly. Data synthesis is a viable alternative and succeeds on closed tasks, yet the humanities and social sciences (HSS) are overlooked, and their open-ended nature makes synthesis challenging. Moving beyond prior capability-centric, fragmented attempts, we adopt a subject-centric paradigm, define the first HSS domain system covering 14 mainstream fields, and introduce HSS-Synth, the first data synthesis pipeline for HSS. HSS-Synth comprises: (1) constructing seed documents from web corpora via multi-step filtering and text refinement evaluated by a judge; (2) specifying "requirements + persona" to backtranslate seed documents into diverse yet faithful instructions with a strict Q&A alignment check; and (3) breaking LLM response limits via teacher-forced Answering that feeds seed documents during response generation to anchor semantics, reduce hallucinations, and preserve tone and integrity. HSS-Synth yields 237k high-quality, diverse instruction-tuning samples that outperform 14 leading baselines on 16 benchmarks. The fine-tuned Qwen3-8B-Base sets a new SOTA and approaches the official Qwen3-8B, improving both human preference and knowledge capabilities without performance seesaws. Extensive experiments demonstrate HSS-Synth's robustness and transferability. Our code is publicly available at https://github.com/pengr/HSS-Synth.
CommentsACL Findings 2026 Paper
