Research Intern - Large Foundational Model (LFM) Evaluation

Overview Research Internships at Microsoft provide a dynamic environment for research careers with a network of world-class research labs led by globally-recognized scientists and engineers, who pursue innovation in a range of scientific and technical disciplines to help solve complex challenges in diverse fields, including computing, healthcare, economics, and the environment. The Special Projects team at the Redmond Lab is offering research internships focused on examining the robustness of large foundational model (LFM) and addressing the challenges associated with model evaluation. Specifically, the research aims to go beyond standard benchmarks to more effectively evaluate various types of large language models (LLMs), from foundational models to those fine-tuned for specific tasks. The goal is to move toward real-world, evidence-based performance guarantees and improved skill transferability for LLMs. Interns will benefit from access to specialized hardware, proprietary models, and real-world evidence (RWE) data. Ideal candidates should possess hands-on experience working with LLMs and be familiar with popular prompt-based methods such as few-shot, chain-of-thought, and graph-of-thought techniques. Experience in instruction-based and explanation-based fine-tuning, as well as knowledge distillation, is highly desirable. Furthermore, candidates should be well-versed in standard evaluation metrics like ROUGE and BERT-Score and have an understanding of graph and embedding methods. The Research Intern will join a multidisciplinary team composed of machine learning researchers, applied statisticians, computer scientists, and AI ethicists. Qualifications Required Qualifications Must be accepted or currently enrolled in a PhD program in Computer Science or related STEM field. Other Requirements Research Interns are expected to be physically located in their manager’s Microsoft worksite location for the duration of their internship.In addition to the qualifications above, you’ll need to submit a minimum of two reference letters for this position. After you submit your application, a request for letters may be sent to your list of references on your behalf. Note that reference letters cannot be requested until after you have submitted your application, and furthermore, that they might not be automatically requested for all candidates. You may wish to alert your letter writers in advance, so they will be ready to submit your letter. Preferred Qualifications Experience with real analysis, functional analysis, and category theory, with applications in machine learning. Experience with embedding space analysis, dimensionality reduction, and embedding visualization. Experience with representation learning or disentangling representations in either text or other modalities. Experience in using Lipschitz continuity to solve various problems in machine learning. Experience in neural network interpretability. The base pay range for this internship is USD $6,290 - $12,170 per month. There is a different range applicable to specific work locations, within the San Francisco Bay area and New York City metropolitan area, and the base pay range for this role in those locations is USD $8,060 - $13,240 per month. Certain roles may be eligible for benefits and other compensation. Find additional benefits and pay information here: Responsibilities Research Interns put inquiry and theory into practice. Alongside fellow doctoral candidates and some of the world’s best researchers, interns learn, collaborate, and network for life. Research Interns not only advance their own careers, but they also contribute to exciting research and development strides. During the 12-week internship, Research Interns are paired with mentors and expected to collaborate with other interns and researchers, present findings, and contribute to the vibrant life of the community. Research internships are available in all areas of research, and are offered year-round, though they typically begin in the summer. #J-18808-Ljbffr Microsoft