Michael R. Douglas (Harvard CMSA)

https://youtu.be/h3jcSg359E8 New Technologies in Mathematics Seminar Speakers: João Araújo, Mathematics Department, Universidade Nova de Lisboa and Michael Kinyon, Department of Mathematics, University of Denver Title: From Engine to Auto Abstract: Bill McCune produced the program EQP that deals with first order logic formulas and in 1996 managed to solve Robbins' Conjecture. This very powerful tool reduces […]

https://youtu.be/g8zdumOAWzw New Technologies in Mathematics Seminar Speaker: Cyril Zhang, Microsoft Research Title: How do Transformers reason? First principles via automata, semigroups, and circuits Abstract: The current "Transformer era" of deep learning is marked by the emergence of combinatorial and algorithmic reasoning capabilities in large sequence models, leading to dramatic advances in natural language understanding, program synthesis, […]

https://youtu.be/ztk5TPYTKZA New Technologies in Mathematics Seminar Speaker: Jimmy Ba, University of Toronto Title: How to steer foundation models? Abstract: By conditioning on natural language instructions, foundation models and large language models (LLMs) have displayed impressive capabilities as general-purpose computers. However, task performance depends significantly on the quality of the prompt used to steer the model. […]

https://youtu.be/UID_oXuN-0Y New Technologies in Mathematics Seminar Speaker: Timo Schick, Meta AI Title: Toolformer: Language Models Can Teach Themselves to Use Tools Abstract: Language models exhibit remarkable abilities to solve new tasks from just a few examples or textual instructions, especially at scale. They also, paradoxically, struggle with basic functionality, such as arithmetic or factual lookup, where […]

https://youtu.be/5LXiQUsnHrI New Technologies in Mathematics Seminar Speaker: Dmitry Krotov, IBM Research - Cambridge Title: Modern Hopfield Networks for Novel Transformer Architectures Abstract: Modern Hopfield Networks or Dense Associative Memories are recurrent neural networks with fixed point attractor states that are described by an energy function. In contrast to conventional Hopfield Networks, which were popular in […]

https://youtu.be/wTQH6mRDXhw New Technologies in Mathematics Seminar Speaker: Ronen Eldan, Microsoft Research Title: The TinyStories Dataset: How Small Can Language Models Be And Still Speak Coherent Abstract: While generative language models exhibit powerful capabilities at large scale, when either the model or the number of training steps is too small, they struggle to produce coherent and fluent […]

https://youtu.be/Sc6k06wVX3s New Technologies in Mathematics Seminar Speaker: François Charton, Meta AI Title:  Transformers for maths, and maths for transformers Abstract: Transformers can be trained to solve problems of mathematics. I present two recent applications, in mathematics and physics: predicting integer sequences, and discovering the properties of scattering amplitudes in a close relative of Quantum ChromoDynamics. Problems of […]

https://youtu.be/u-pkmdkQoMU New Technologies in Mathematics Seminar Speaker: Alex Gu, MIT Dept. of EE&CS Title: LeanDojo: Theorem Proving with Retrieval-Augmented Language Models Abstract: Large language models (LLMs) have shown promise in proving formal theorems using proof assistants such as Lean. However, existing methods are difficult to reproduce or build on, due to private code, data, and large compute […]

https://youtu.be/M25cbX5do8Y New Technologies in Mathematics Seminar Speaker: Yuanzhi Li, CMU Dept. of Machine Learning and Microsoft Research Title: Physics of Language Models: Knowledge Storage, Extraction, and Manipulation Abstract: Large language models (LLMs) can memorize a massive amount of knowledge during pre-training, but can they effectively use this knowledge at inference time? In this work, we show several striking […]

https://youtu.be/bRHb-MVExJ4 New Technologies in Mathematics Seminar Speaker: Sean Welleck, CMU, Language Technologies Institute Title: Llemma: an open language model for mathematics Abstract: We present Llemma: 7 billion and 34 billion parameter language models for mathematics. The Llemma models are initialized with Code Llama weights, then trained on the Proof-Pile II, a 55 billion token dataset of […]