BEGIN:VCALENDAR VERSION:2.0 PRODID:researchseminars.org CALSCALE:GREGORIAN X-WR-CALNAME:researchseminars.org BEGIN:VEVENT SUMMARY:Sergei Gukov (Caltech) DTSTART:20240929T070000Z DTEND:20240929T081000Z DTSTAMP:20260404T140809Z UID:AI-Math-2024/1 DESCRIPTION:Title: Mathematics and AI: latest trends and future\nby Sergei Gu kov (Caltech) as part of AI and Mathematics: Current Trends and Future Dir ections\n\nLecture held in Bar-Ilan University.\n\nAbstract\nIn this talk\ , intended for a broad audience\, I will use concrete examples from combin atorial group theory and low-dimensional topology to illustrate how rapid growth of AI algorithms can change the way we do mathematical research and help us with some of the most difficult mathematical challenges. One of t he goals of this talk is to provide a gentle introduction to some of the m odern tools in Machine Learning\, in part explaining its increasing role i n everyday life and in pure mathematics as well.\n LOCATION:https://stable.researchseminars.org/talk/AI-Math-2024/1/ END:VEVENT BEGIN:VEVENT SUMMARY:Boris Yangel (Nebius) DTSTART:20240929T081500Z DTEND:20240929T090000Z DTSTAMP:20260404T140809Z UID:AI-Math-2024/2 DESCRIPTION:Title: Teaching LLM agents to automate software engineering tasks \nby Boris Yangel (Nebius) as part of AI and Mathematics: Current Trends a nd Future Directions\n\nLecture held in Bar-Ilan University.\n\nAbstract\n LLM technology has come to a point where it becomes possible to build agen tic systems that can perform complex sequences of actions over a code repo sitory to implement new features and fix bugs in an interactive environmen t\, provided with just a textual description of the issue that needs to be resolved. In this talk I will give an overview of this emergent avenue of research\, present our recent results regarding training open source mode ls to perform well in this domain\, and talk about the challenges that we have faced.\n LOCATION:https://stable.researchseminars.org/talk/AI-Math-2024/2/ END:VEVENT BEGIN:VEVENT SUMMARY:Nachum Dershowitz (Tel Aviv University) DTSTART:20240929T101000Z DTEND:20240929T104000Z DTSTAMP:20260404T140809Z UID:AI-Math-2024/3 DESCRIPTION:Title: Language Models have a Hard Time Thinking Logically\nby Na chum Dershowitz (Tel Aviv University) as part of AI and Mathematics: Curre nt Trends and Future Directions\n\nLecture held in Bar-Ilan University.\n\ nAbstract\nI will illustrate some of the difficulties modern generative la nguage models have with logical reasoning (propositional and first order) and with self reflection.\n LOCATION:https://stable.researchseminars.org/talk/AI-Math-2024/3/ END:VEVENT BEGIN:VEVENT SUMMARY:Yuval Dolev (Bar-Ilan University) DTSTART:20240929T104500Z DTEND:20240929T113000Z DTSTAMP:20260404T140809Z UID:AI-Math-2024/4 DESCRIPTION:Title: AI\, creativity\, and fundamental questions in the philosophy of mathematics\nby Yuval Dolev (Bar-Ilan University) as part of AI and Mathematics: Current Trends and Future Directions\n\nLecture held in Bar- Ilan University.\n\nAbstract\nCan AI be expected to be mathematically crea tive? I offer reasons why the answer should be negative. I do so in the co ntext of two central doctrines in the philosophy of mathematics – Platon ism and intuitionism. I argue that despite being opposed to each other on almost any question regarding the nature of mathematical objects and truth \, these two rival positions do share this – that creativity is not redu cible either to randomness\, or to a deterministic computation. Moreover\, I suggest there is a normative (aesthetic) foundation to mathematics\, to which AI is blind. These considerations highlight the limitedness of AI a nd point to the indispensability and irreplaceability of human creativity. \n LOCATION:https://stable.researchseminars.org/talk/AI-Math-2024/4/ END:VEVENT BEGIN:VEVENT SUMMARY:Kevin Buzzard (Imperial College London) DTSTART:20240929T120000Z DTEND:20240929T123000Z DTSTAMP:20260404T140809Z UID:AI-Math-2024/5 DESCRIPTION:Title: Formalising Fermat\nby Kevin Buzzard (Imperial College Lon don) as part of AI and Mathematics: Current Trends and Future Directions\n \nLecture held in Bar-Ilan University.\n\nAbstract\nRight now it seems tha t humanity has got to the stage where teaching an interactive theorem prov er a proof of Fermat's Last Theorem will be technically possible\, but a l ot of work. What actually is involved\, and how can AI help?\n LOCATION:https://stable.researchseminars.org/talk/AI-Math-2024/5/ END:VEVENT BEGIN:VEVENT SUMMARY:Amit Somech (Bar-Ilan University) DTSTART:20240929T123500Z DTEND:20240929T130500Z DTSTAMP:20260404T140809Z UID:AI-Math-2024/6 DESCRIPTION:Title: Data-Speaking LLM Agents\nby Amit Somech (Bar-Ilan Univers ity) as part of AI and Mathematics: Current Trends and Future Directions\n \nLecture held in Bar-Ilan University.\n\nAbstract\nWe all know that LLMs are great for all kinds of everyday tasks—writing\, editing\, summarizin g articles\, and even generating code on demand. But can we rely on them f or data-focused tasks? On one hand\, LLMs are trained on massive text data sets\, some of which include structured data like CSVs and JSON files. Plu s\, thanks to forums like Stack Overflow\, they should be able to help wit h writing SQL queries and data-processing code. However\, we’ve found th at when it comes to working with data\, LLMs aren't as "fluent" as we migh t hope. When faced with larger tables\, they often struggle to fully under stand the structure or produce accurate tabular output. They also aren’t great at verifying facts when the information is presented in table form\ , even if they "know" the facts. In this talk\, we’ll introduce several non-trivial data tasks\, review how standard LLMs perform\, and explore ne w architectures that could push the current capabilities forward.\n LOCATION:https://stable.researchseminars.org/talk/AI-Math-2024/6/ END:VEVENT BEGIN:VEVENT SUMMARY:Alexei Miasnikov (Stevens Institute of Technology) DTSTART:20240929T131500Z DTEND:20240929T140000Z DTSTAMP:20260404T140809Z UID:AI-Math-2024/7 DESCRIPTION:Title: Revolutionizing Math Education: Harnessing AI and Algorithms f or Student Success\nby Alexei Miasnikov (Stevens Institute of Technolo gy) as part of AI and Mathematics: Current Trends and Future Directions\n\ nLecture held in Bar-Ilan University.\n\nAbstract\nWhile AI has captured w idespread attention for its applications in various fields\, its role in m athematics education presents both exciting opportunities and distinct cha llenges. This talk will focus on Gradarius\, a system that uses graph algo rithms to provide step-by-step feedback on student solutions\, offering un matched precision and rigor. We will explore how AI is integrated into thi s framework—not as the core mathematical engine\, but as a tool to enhan ce usability\, analyze student behavior\, and adapt to diverse learning pa ths. This presentation will offer a balanced view of AI’s strengths and limitations\, with a focus on how cutting-edge algorithms and machine lear ning can work together to revolutionize the teaching and learning of mathe matics.\n LOCATION:https://stable.researchseminars.org/talk/AI-Math-2024/7/ END:VEVENT END:VCALENDAR