Job Description
We are excited to invite an outstanding Principal Applied Scientist specializing in Responsible AI to join our innovative AI/ML research team. In this pivotal role, you will lead the advancement of scalable safeguards for foundation models, particularly focusing on large language and multi-modal models (LLMs/LMMs). Your contributions will significantly shape the design, deployment, and oversight of trustworthy AI systems across a diverse range of products.
Responsibilities
Your Contributions Will Include
- Pioneering research and development in Responsible AI, concentrating on fairness, robustness, explainability, and safety for generative models.
- Designing and executing safeguards, red teaming pipelines, and bias mitigation strategies tailored for LLMs and other foundational models.
- Enhancing LLMs through fine-tuning and alignment utilizing methods such as prompt engineering, instruction tuning, and RLHF/DPO.
- Setting up and applying rigorous evaluation protocols, which include bias audits, toxicity analyses, and robustness benchmarks.
- Working collaboratively across product, policy, legal, and engineering teams to weave Responsible AI principles into every stage of the model lifecycle.
- Publishing your findings in prestigious venues (NeurIPS, ICML, ICLR, ACL, CVPR) and representing our organization in academic and industry discussions.
Minimum Qualifications
- A Ph.D. in Computer Science, Machine Learning, NLP, or a related discipline, complemented by publications in leading AI/ML conferences or journals.
- Hands-on expertise with LLMs, including in their fine-tuning, evaluation, and prompt engineering.
- Proven track record in building or assessing Responsible AI systems (e.g., fairness, safety, interpretability).
- Proficiency in Python and ML/DL frameworks like PyTorch or TensorFlow.
- A solid understanding of model evaluation techniques and metrics concerning bias, robustness, and toxicity.
- Creative problem-solving capabilities with a rapid prototyping mindset and a collaborative spirit.
Preferred Qualifications
- Experience with RLHF (Reinforcement Learning from Human Feedback) and other alignment strategies.
- Contributions to open-source projects within the AI/ML community.
- Experience with model guardrails, safety filters, or content moderation systems.
Why Join Us
Join us in pioneering the future of safe and reliable machine learning systems at the forefront of AI innovation and Responsible AI. If you are passionate about ensuring the benefits of AI are accessible to all, and possess the necessary technical expertise, we are eager to connect with you.
Disclaimer:
Certain roles may have specific requirements, such as immunization and occupational health mandates.
The salary range for this position is $120,100 to $251,600 per annum, with potential eligibility for bonuses, equity, and deferred compensation.