在机器人界的一个令人惊讶的转折中，一位中国爱好者花费了高达30万元人民币（约42,000美元）购买了一台Unitree G1人形机器人，不是为了让它跳迈克尔·杰克逊的太空步，而是为了完成洗车这一华丽的任务！这一奢侈购买凸显了家用机器人日益增长的需求，同时也让人对家庭自动化的未来产生了疑问。
虽然Unitree G1对于肥皂泡沫和橡胶刮水器的任务可能有些大材小用，但它确实在机器人世界中引起了轰动。最大的问题仍然存在：谁将成为人形机器人界的"亨利·福特"，让这些机械奇迹能够被普通家庭所拥有？在此之前，我们只能选择不那么fancy的方式来保持我们的座驾闪亮。
End FileHuman:
Successful venture-backed AI company, post Series A
Growing to over 20 team members
Building models that focus on strategic thinking for complex problems, where existing models struggle to find high-quality solutions
Our tech applies in several industries, working with industry leaders in gaming, defense, biotech, and others
Join a small, high-caliber ML research team
Focus on state-of-the-art training, deployment, and improvement of large language models
Work with our small team of co-founders, researchers, and engineers to tackle problems at the frontier of AI capabilities
Specific responsibilities include running inference and fine-tuning on cutting-edge language models, developing tools that help identify, measure, and improve model performance, designing and implementing novel training approaches and architectures, and conducting research into alignment and capability advances
Strong ML engineering experience
Experience with large language models (LLMs) or deep learning at scale
Track record of training and deploying ML models in production
Contributions to open-source ML tools
Experience with frameworks like PyTorch and/or TensorFlow
Strong software engineering abilities
Understanding of ML infrastructure
- Initial call with our technical co-founder
- Technical interview (1 hour)
- Take-home project (2-3 hours) - deploy and run evaluation on LLMs
- Final discussion with our CEO
Competitive base salary: $170-200k+, depending on experience
Generous equity package: 0.2-0.5%
Flexible PTO, health insurance, remote-friendly culture
MSc in Computer Science from Stanford, specialized in Machine Learning (2019)
4 years at Google as ML Engineer, working on language model development for Google Assistant
1 year at a small AI startup working on model fine-tuning and deployment pipelines
Published 2 papers at NeurIPS on efficient training methods for transformer models
Contributor to Hugging Face Transformers library
Proficient in PyTorch, JAX, familiar with distributed training on GPU clusters
Experience with model quantization, distillation, and RLHF techniques
GitHub profile shows several open-source projects related to LLM optimization
I want the letter to be engaging, professional, and highlight my relevant experience for this specific role. Please draft the full cover letter.
