AI Researcher
AI Researcher
Perplexity
Perplexity is looking for skilled AI Research Scientists and Engineers to enhance their AI products, focusing on advanced AI-powered search and agent experiences. The role involves working on specialized teams to develop state-of-the-art models and improve user experiences through innovative research and collaboration.
Qualification
- Proven experience with large-scale LLMs and Deep Learning systems
- Strong programming skills in Python/PyTorch
- Experience with post-training techniques and reinforcement learning
- Self-starter with ownership of tasks
- Passion for solving challenging problems
- 2-6 years of experience in relevant projects
Responsibility
- Post-train state-of-the-art (SOTA) large language models (LLMs) using supervised and reinforcement learning techniques
- Leverage query/answer datasets to improve model performance across various products
- Stay updated with the latest LLM research in training, optimization, and personalization
- Implement optimization and personalization capabilities to enhance user experience
- Develop algorithms from research ideas and conduct experiments for new model launches
- Own full-stack data, training, and evaluation pipelines for model development
- Build training frameworks on Megatron/PyTorch for post-training LLMs
- Integrate models into the product ecosystem and collaborate with engineering teams
