Eugene Yan
PulseAugur coverage of Eugene Yan — every cluster mentioning Eugene Yan across labs, papers, and developer communities, ranked by signal.
-
Eugene Yan: AI collaboration requires context, configuration, and memory
Eugene Yan outlines a framework for effectively collaborating with AI tools, emphasizing practices that enable compounding improvements over time. Key principles include providing clear context, encoding personal prefer…
-
Eugene Yan outlines a 3-step process for effective LLM product evaluations
Eugene Yan's guide outlines a three-step process for developing product evaluations for LLMs. The first step involves labeling a small dataset, focusing on binary pass/fail or win/lose labels to ensure clarity and consi…
-
Eugene Yan offers advice for new principal tech engineers and scientists
Eugene Yan's advice for principal-level technical individual contributors emphasizes a shift from individual coding to broader influence and technical vision. Principals are encouraged to be hands-on but focus their cor…
-
Eugene Yan trains LLM-recommender hybrid for steerable, explainable recommendations
Eugene Yan has developed a novel approach to recommender systems by training a hybrid language model that understands both natural language and item IDs. This model, which extends the vocabulary of a language model with…
-
Eugene Yan 探讨卓越领导力特质：愿景、执行和同理心。
Eugene Yan 的文章概述了卓越的领导力特质，将其分为愿景、执行和同理心。卓越的领导者兼具这三者，而优秀的领导者至少擅长其中两者。文章进一步详细阐述了领导者如何提供灵感和资源，扫除障碍，并通过放手来赋能团队。文章还触及了适用于模糊、高风险环境（突击队式）和快速增长阶段（士兵式）的领导风格。
-
Eugene Yan 使用 Amazon Q 和 MCP 构建新闻代理以进行每日回顾
Eugene Yan 开发了一个使用由 Amazon Q CLI 和 MCP 驱动的代理工作流生成每日新闻回顾的系统。该系统将新闻源分割成块，并由单独的子代理处理每个块。主代理随后将子代理的摘要整合为最终的每日回顾，展示了基于代理的新闻聚合的实际应用。
-
Eugene Yan: LLM-as-judge won't fix AI product evals; focus on process
Eugene Yan argues that relying solely on tools like LLM-as-judge will not fix product evaluation issues. Instead, he emphasizes that a robust evaluation process, akin to the scientific method, is crucial for improving A…
-
Eugene Yan discusses building LLM-powered applications at NVIDIA GTC 2025
Eugene Yan presented at NVIDIA GTC 2025 on a panel discussing the development of applications powered by large language models. The session, titled "Insights and Lessons Learned From Building LLM-Powered Applications," …
-
Eugene Yan explores the paradoxical rules of effective writing
Eugene Yan's article explores the paradoxical nature of writing, suggesting that effective writing often involves embracing seemingly contradictory approaches. He posits that clarity can be achieved through both simple …
-
技术作者分享通过AI内容建立受众的策略
技术作者 Hamel Husain 分享了建立受众的策略，强调与他人作品进行真实互动和持续内容创作。他建议开发者为现有讨论增加价值，并强调刻意练习和提高文案写作技巧的重要性。Husain 还建议利用 AI 工具来简化内容创作并建立一个从语音到内容的管道。
-
Eugene Yan 分享举办每周 AI 论文俱乐部以建立学习社区的指南
Eugene Yan 详细介绍了其成功的每周论文俱乐部，该俱乐部已运行 18 个月，讨论了至少 80 篇与 AI 相关的论文。俱乐部专注于机器学习中的基础概念、模型、训练和推理技术。Yan 为他人建立类似的学习社区提供了实用指南，强调了持续的日程安排、预读和引导式讨论，以促进技术理解和建立专业人脉。
-
Eugene Yan shares minimal MacBook Pro setup guide for developers
Eugene Yan details his minimal setup for a new M4 MacBook Pro, emphasizing a clean slate approach over restoring from a backup. He outlines configurations for macOS settings, essential developer tools like Homebrew, War…
-
Weights & Biases Hackathon 展示了富有创意的 LLM 评估项目
Weights & Biases LLM-Evaluator Hackathon 的评委 Eugene Yan 分享了此次活动的见解，超过 100 名参与者构建了创意项目。团队专注于知识图谱构建、个性特征的 LLM 评估以及提示优化等领域。Yan 讨论了使用 LLM 评估器的关键考虑因素，包括评分方法和性能指标，并对团队在周末取得的快速进展印象深刻。
-
How To Hire AI Engineers — with James Brady & Adam Wiggins of Elicit
Hiring managers for AI and Machine Learning roles should focus on a structured interview process that assesses both technical and non-technical skills. Key areas include software engineering proficiency, demonstrated th…
-
Eugene Yan shares lessons learned from a year of building with LLMs
Eugene Yan's article distills a year's worth of experience in developing applications powered by large language models. The insights cover a broad spectrum, from the practical, hands-on aspects of implementation to the …
-
Eugene Yan 推出 AlignEval 以简化和自动化 LLM 评估
Eugene Yan 推出了 AlignEval，这是一款旨在简化和自动化大型语言模型 (LLM) 评估过程的新应用程序。该工具引导用户上传数据、将样本标记为通过或失败、定义评估标准以及优化基于 LLM 的评估器。AlignEval 强调数据优先的方法，鼓励用户从实际模型输出来推导评估标准，而不是预定义的指标，旨在减少 AI 产品开发中的瓶颈。
-
Eugene Yan 建议不要在单元测试中模拟机器学习模型
Eugene Yan 的文章讨论了将传统单元测试实践应用于机器学习代码的挑战。与手工编写逻辑的标准软件不同，ML 模型从数据中学习逻辑，使得直接测试这种学习到的逻辑变得复杂。Yan 建议，虽然在软件中模拟依赖项很常见，但 ML 单元测试可能需要与实际模型进行交互，特别是为了验证训练进度或推理的正确性。他提出使用小型、自包含的数据样本，并使用随机或空权重进行测试，以克服大型模型尺寸和推理速度慢的问题。
-
AI models can now be fine-tuned using synthetic data, reducing costs and privacy risks
Synthetic data, generated by models or simulations rather than real-world sources, offers a faster and more cost-effective alternative to human annotation for fine-tuning AI models. This approach can lead to improved mo…
-
Eugene Yan curates essential language modeling papers for study groups
Eugene Yan has compiled a reading list of fundamental language modeling papers, intended to facilitate group study sessions. The list includes seminal works like "Attention Is All You Need," "BERT," and "GPT-3," each ac…
-
Eugene Yan explores effective push notification strategies for engagement and relevance
Push notifications can be viewed as a proactive recommender system, but they present unique challenges compared to traditional recommendations. Unlike search or on-site suggestions, push notifications lack clear user in…