Suggestions
Anima Anandkumar
AI Pioneer, AI+Science, Professor at Caltech, Former Senior Director of AI at NVIDIA, Former Principal Scientist at AWS AI.
Anima Anandkumar is a prominent figure in the fields of machine learning and artificial intelligence, currently serving as the Bren Professor of Computing and Mathematical Sciences at the California Institute of Technology (Caltech). She also holds the position of Director of Machine Learning Research at NVIDIA, where she leads efforts in developing advanced AI algorithms.
Educational Background
- B.Tech in Electrical Engineering: Indian Institute of Technology (IIT) Madras, 2004
- PhD in Electrical Engineering: Cornell University, 2009
- Postdoctoral Research: MIT, 2009-2010
Academic and Professional Career
Anandkumar began her academic career as an assistant professor at the University of California, Irvine (UCI) in 2010, where she progressed to associate professor before moving to Caltech in 2018. During her tenure at UCI, she worked on tensor decompositions and large-scale machine learning. She has also held positions at Amazon Web Services (AWS) as a principal scientist and has been involved with Microsoft Research.
Research Contributions
Anandkumar's research focuses on:
- Tensor Algebra: Developing tensor algorithms that enhance the processing of multidimensional data.
- Deep Learning: Exploring non-convex optimization problems and their applications in AI.
- Neural Operators: Innovating methods that extend deep learning capabilities to model complex scientific processes.
Her work has led to significant advancements in various applications including weather forecasting, drug discovery, and engineering design. Notably, she was awarded the ACM Gordon Bell Special Prize for her contributions to high-performance computing during the COVID-19 pandemic.
Awards and Recognition
Anandkumar has received numerous accolades throughout her career:
- Alfred P. Sloan Fellowship
- NSF Career Award
- Recognitions from organizations such as Microsoft, Google, and Adobe.
- She was named a 2023 Guggenheim Fellow and is part of the World Economic Forum’s Expert Network.
Initiatives at Caltech
At Caltech, Anandkumar co-founded the AI4Science initiative, which aims to leverage AI for scientific research. She is also involved with various advisory roles and has spoken at numerous prestigious conferences.
Anima Anandkumar's contributions to AI and machine learning continue to influence both academic research and practical applications across multiple domains.
Highlights
Honored to be a Blavatnik Family Foundation Award finalist. I want to thank my Caltech team and collaborators. https://lnkd.in/gc2iEgVy
This award represents the team's efforts. We have been working in AI+Science before it went mainstream. Our Neural Operators have revolutionized scientific modeling, including creating the first AI-based weather model, inventing better medical devices, and modeling fusion reactions, among many others.
Congratulations Jiawei Zhao on an excellent PhD defense! Jiawei has been a pioneer in hardware-efficient training. Slides https://lnkd.in/gwSzdQQY When he started his PhD, everyone was focusing on inference efficiency, and training runs were small, Jiawei took the bold step to pursue training efficiency. Jiawei's latest work Galore has led to huge reduction (>80%) of memory requirements for pretraining LLMs. It is based on a principled approach of projecting the gradient to a low-rank subspace while keeping the weights as full rank. Jiawei also worked on reduced precision training in logarithmic number system (LNS). The LNS is extreme quantization: instead of a standard floating-point system with exponent and mantissa, LNS only has the exponent part. Jiawei proposed a multiplicative update method to directly train in LNS. This will impact training on smaller edge devices: which is not yet mainstream today. Jiawei first started working on signSGD: surprisingly just communicating sign of gradients in distributed training is sufficient for accuracy while significantly reducing bandwidth requirements. A variant of it (Lion optimizer) is now used for LLMs. Jiawei has also done in-depth work on initializing neural networks with just zeros and ones, instead of random initialization. He has shown that this can help with better reproducibility while preserving accuracy. Congratulations Jiawei Zhao
