Machine Learning Research Scientist – New Grad – Hugging Face
Hugging Face is the AI community's home — the open-source platform hosting 900,000+ pre-trained models, 200,000+ datasets, and 300,000+ ML applications used by 5+ million developers and researchers worldwide. Hugging Face's Transformers library is the most popular ML library on GitHub with 130,000+ stars, and the company runs the largest collaborative AI research platform in the world. From BERT and GPT-2 to the latest Llama and Mistral models, every major AI breakthrough starts as open weights on Hugging Face. We are hiring New Grad ML Research Scientists to advance the science of open-source AI.
Responsibilities
- Conduct research on large language model capabilities — evaluating pre-trained models, developing new benchmarks, and publishing findings on Hugging Face's Open LLM Leaderboard
- Build and maintain the Hugging Face Evaluate library — adding new metrics, benchmarks, and evaluation datasets for the ML research community
- Develop datasets and fine-tuning experiments using the Hugging Face TRL (Transformer Reinforcement Learning) library for instruction tuning and RLHF alignment
- Contribute to Hugging Face's open-source research papers on efficient training, model compression, quantization, and inference optimization
- Analyze model behavioral differences across architectures, scales, and training data to identify capability gaps and alignment failures
- Engage with Hugging Face's 5M+ developer community through research blog posts, model cards, and dataset documentation
Requirements
- Bachelor's or Master's degree in Machine Learning, Computer Science, or Statistics
- Hands-on experience with Hugging Face Transformers, PyTorch, or JAX for model training and evaluation
- Strong Python skills and comfort working in open-source research environments
- Genuine passion for open-source AI and democratizing access to state-of-the-art ML models
- Research publications, competitive ML submissions, or significant open-source contributions are a strong plus
Benefits
- Competitive salary with Hugging Face equity
- Hugging Face Pro Hub access and GPU compute credits for research
- Medical, dental, and vision benefits
- 401(k) with Hugging Face contribution
- Remote-first culture with the world's most vibrant open-source AI community