1
AI data startup Micro1 reaches $500M gross run rate amid AI training boom
AI数据公司Micro1营收狂奔至5亿美元，揭示合成数据与训练需求爆发。
Surging demand for AI training data is driving rapid growth for the startup and its rivals.
2
图灵奖得主萨顿警告：AI 行业过度依赖合成数据是一个大错误
图灵奖得主萨顿泼冷水：合成数据模拟不了真实世界的复杂，AI巨头该冷静了。
IT之家 8 月 19 日消息，里奇 · 萨顿（Rich Sutton）是当今人工智能热潮背后关键技术的先驱之一，但如今，他认为科技巨头为了让 AI 持续发展而采取的路线存在严重问题。 这位加拿大计算机科学家、图灵奖得主在当地时间周二发布的红杉资本播客节目中表示，AI 行业越来越依赖合成训练数据，这…
3
Self-Bootstrapping Automated Program Repair: Using LLMs to Generate and Evaluate Synthetic Training Data for Bug Repair
用LLM合成训练数据，破解自动修Bug的数据瓶颈，思路新颖且实战价值高。
arXiv:2505.07372v3 Announce Type: replace-cross Abstract: This paper presents a novel methodology for enhancing Automated Program Repair (APR) through…
4
LLMs Get Smarter from Targeted Synthetic Multilingual Data
针对多语言提示下大模型表现不一的难题，用定向合成数据提升模型智能，值得关注。
arXiv:2608.15964v1 Announce Type: cross Abstract: Language-specific competency (LSC) is the phenomenon of a language model performing better or worse …
5
Constraint-Aware Synthetic Tabular Data Generation via Inter-Column Constraint Discovery with LLM Agents
用LLM智能体自动发现列间约束，让合成表格数据更真实可信，为数据增强与隐私保护提供新思路。
arXiv:2608.15109v1 Announce Type: new Abstract: Generating structurally valid synthetic tabular data remains difficult: outputs with high statistical …
6
Validated Synthetic Patient Generation for Small Longitudinal Cohorts: Coagulation Dynamics Across Pregnancy
用AI合成患者数据破解小样本难题，精准还原妊娠凝血动态变化，医学研究新利器。
arXiv:2604.07557v2 Announce Type: replace Abstract: Small longitudinal cohorts, common in maternal health, rare diseases, and early-phase trials, limi…
7
Clinical Communication Processing with Models Trained on LLM-Generated Synthetic Data: A Structured Survey and Novel Application Case Studies
临床通信遇上AI：用大模型合成数据训练，破解医疗文本处理的数据难题，技术路径与实战案例一次讲透。
arXiv:2608.05993v1 Announce Type: new Abstract: Much clinical value is conveyed not through structured records but through communication: exchanges in…
8
Leveraging Synthetic Data for Question Answering with Multilingual LLMs in the Agricultural Domain
用合成数据驯服农业领域多语言大模型，破解低资源语言问答难题，实验严谨、方法可复用。
arXiv:2507.16974v3 Announce Type: replace-cross Abstract: Enabling farmers to access accurate agriculture-related information in their native language…
9
FineInstructions: Scaling Synthetic Instructions to Pre-Training Scale
通过合成指令数据扩展预训练规模，突破传统监督训练数据瓶颈的新方法
arXiv:2601.22146v2 Announce Type: replace-cross Abstract: Due to limited supervised training data, large language models (LLMs) are typically pre-trai…
10
36氪首发 | 前博世自动驾驶算法工程师创业，用合成数据做触觉大模型
触觉手套（图源/企业） 本文约 2400 字，建议阅读 6 分钟 作者丨欧雪 编辑丨袁斯来 硬氪获悉，空间智能公司「大衍科技」近期完成数千万元天使轮融资，由松禾资本领投，浙江省省金控与广州番禺创新基金等国资背景机构参与。资金将主要用于触觉大模型研发、机器人数据产线建设及团队扩张。 「大衍科技」202…
11
Rectification Difficulty and Optimal Sample Allocation in LLM-Augmented Surveys
利用大语言模型低成本生成调查数据时，准确性因问题波动，本文研究如何最优分配固定的人类样本预算以提升估计效果。
arXiv:2604.17267v2 Announce Type: replace Abstract: Large Language Models can generate synthetic survey responses at low cost, but their accuracy vari…
12
Structured Prompting and Automated Evaluation in Fixed Synthetic Japanese-Language Counseling Dialogues
用结构化提示提升合成日语咨询对话质量，并探索自动评估方法，为对话系统研究提供新思路。
arXiv:2507.02950v3 Announce Type: replace-cross Abstract: Large language models (LLMs) may support counseling training, yet evidence from Japanese-lan…
13
TRACE: Capability-Targeted Agentic Training
针对智能体任务失败根因，提出按能力缺口定向训练的新范式，为Agent训练提供精准解法。
arXiv:2604.05336v2 Announce Type: replace Abstract: Models often fail to complete agentic tasks because they lack core capabilities required by the ta…
14
Libra: Training the Environment for Agentic Information Retrieval
揭秘智能体信息检索新范式：将数据驱动优化从模型转向工作环境本身
arXiv:2607.00016v1 Announce Type: cross Abstract: Information localization within massive repositories is a cornerstone of agentic LLM systems. While …
15
Self-Study Reconsidered: The Hidden Fragility of Learning from Self-Generated QA
揭示语言模型自我生成QA训练的隐藏脆弱性，一篇值得关注的AI研究论文。
arXiv:2606.32002v1 Announce Type: new Abstract: Language models are increasingly taught from synthetic question--answer (QA) supervision: a model gene…
16
Synthetic Interaction Data for Scalable Personalization in Large Language Models
用合成交互数据破解LLM个性化扩展难题，数据瓶颈的新解法。
arXiv:2602.12394v2 Announce Type: replace Abstract: Personalized prompting offers large opportunities for deploying large language models (LLMs) to di…
17
DocArena: Turning Raw Documents into Controllable Training Environments for Document Search Agents
无需人工标注，DocArena把原始文档变成可控训练场，让文档搜索代理练出更强检索能力，值得关注。
arXiv:2606.26122v1 Announce Type: new Abstract: Recent methods train search agents via reinforcement learning from (question, answer, evidence) tuples…
18
CuratorKIT : Data Curation and Synthetic Data Generation for LLM Post-Training
开源工具助你高效清洗与生成合成数据，优化大模型后训练流程。
arXiv:2606.21631v1 Announce Type: cross Abstract: Data curation is a critical part of post-training pipelines for large language models, yet existing …
19
Evaluating Synthetic Data Generation for Domain Generalization in Fetal Brain MRI Segmentation
评估合成数据在胎儿脑MRI分割域泛化中的效果，探索提升模型跨域鲁棒性的新思路。
arXiv:2411.06842v3 Announce Type: replace-cross Abstract: Fetal brain tissue segmentation from magnetic resonance imaging (MRI) is crucial for studyin…
20
MedSynth: Realistic, Synthetic Medical Dialogue-Note Pairs
用AI生成逼真医疗对话与病历对，解决数据隐私难题，医学NLP新突破。
arXiv:2508.01401v2 Announce Type: replace-cross Abstract: Physicians spend significant time documenting clinical encounters, a burden that contributes…