[47] Niloofar Mireshghallah - Auditing and Mitigating Safety Risks in Large Language Models

[48] Tianqi Chen - Scalable and Intelligent Learning Systems

Tianqi Chen is an Assistant Professor in the Machine Learning Department and Computer Science Department at Carnegie Mellon University and the Chief Technologist of OctoML. His research focuses on the intersection of machine learning and systems. Tianqi's PhD thesis is titled "Scalable and Intelligent Learning Systems," which he completed in 2019 at the University of Washington. We discuss his influential work on machine learning systems, starting with the development of XGBoost,an optimized distributed gradient boosting library that has had an enormous impact in the field. We also cover his contributions to deep learning frameworks like MXNet and machine learning compilation with TVM, and connect these to modern generative AI. - Episode notes: www.wellecks.com/thesisreview/episode48.html - Follow the Thesis Review (@thesisreview) and Sean Welleck (@wellecks) on Twitter - Follow Tianqi Chen on Twitter (@tqchenml) - Support The Thesis Review at www.patreon.com/thesisreview or www.buymeacoffee.com/thesisreview

[47] Niloofar Mireshghallah - Auditing and Mitigating Safety Risks in Large Language Models

Niloofar Mireshghallah is a postdoctoral scholar at the University of Washington. Her research focuses on privacy, natural language processing, and the societal implications of machine learning. Niloofar completed her PhD in 2023 at UC San Diego, where she was advised by Taylor Berg-Kirkpatrick. Her PhD thesis is titled "Auditing and Mitigating Safety Risks in Large Language Models." We discuss her journey into research and her work on privacy and LLMs, including how privacy is defined, common attacks and mitigations, differential privacy, and the balance between memorization and generalization. - Episode notes: www.wellecks.com/thesisreview/episode47.html - Follow the Thesis Review (@thesisreview) and Sean Welleck (@wellecks) on Twitter - Support The Thesis Review at www.patreon.com/thesisreview or www.buymeacoffee.com/thesisreview

[46] Yulia Tsvetkov - Linguistic Knowledge in Data-Driven NLP

Yulia Tsvetkov is a Professor in the Allen School of Computer Science & Engineering at the University of Washington. Her research focuses on multilingual NLP, NLP for social good, and language generation. Yulia's PhD thesis is titled "Linguistic Knowledge in Data-Driven Natural Language Processing", which she completed in 2016 at CMU. We discuss getting started in research, then move to Yulia's work in the thesis that combines ideas from linguistics and natural language processing. We discuss low-resource and multilingual NLP, large language models, and great advice about research and beyond. - Episode notes: www.wellecks.com/thesisreview/episode46.html - Follow the Thesis Review (@thesisreview) and Sean Welleck (@wellecks) on Twitter - Find out more info about the show at www.wellecks.com/thesisreview - Support The Thesis Review at www.patreon.com/thesisreview or www.buymeacoffee.com/thesisreview

[45] Luke Zettlemoyer - Learning to Map Sentences to Logical Form

Luke Zettlemoyer is a Professor at the University of Washington and Research Scientist at Meta. His work spans machine learning and NLP, including foundational work in large-scale self-supervised pretraining of language models. Luke's PhD thesis is titled "Learning to Map Sentences to Logical Form", which he completed in 2009 at MIT. We talk about his PhD work, the path to the foundational Elmo paper, and various topics related to large language models. - Episode notes: www.wellecks.com/thesisreview/episode45.html - Follow the Thesis Review (@thesisreview) and Sean Welleck (@wellecks) on Twitter - Find out more info about the show at www.wellecks.com/thesisreview - Support The Thesis Review at www.patreon.com/thesisreview or www.buymeacoffee.com/thesisreview

[44] Hady Elsahar - NLG from Structured Knowledge Bases (& Controlling LMs)

Hady Elsahar is a Research Scientist at Naver Labs Europe. His research focuses on Neural Language Generation under constrained and controlled conditions. Hady's PhD was on interactions between Natural Language and Structured Knowledge bases for Data2Text Generation and Relation Extraction & Discovery, which he completed in 2019 at the Université de Lyon. We talk about his phd work and how it led to interests in multilingual and low-resource in NLP, as well as controlled generation. We dive deeper in controlling language models, including his interesting work on distributional control and energy-based models. - Episode notes: www.wellecks.com/thesisreview/episode44.html - Follow the Thesis Review (@thesisreview) and Sean Welleck (@wellecks) on Twitter - Find out more info about the show at www.wellecks.com/thesisreview - Support The Thesis Review at www.patreon.com/thesisreview or www.buymeacoffee.com/thesisreview