⏩ LLM Problem Solvers - Limitations of Mathematical Reasoning in LLMs

​we will be looking at:

​https://arxiv.org/pdf/2410.05229

​Understanding the Limitations of Mathematical Reasoning in Large Language Models

​About the paper: There has been a lot of talk about if Language Models can truly reason, recently Apple decided to tackle this question by testing language models ability to reason in previously unseen Mathematical problems. Short Bio of the Discussion Lead: Josh

​Short Bio of the Discussion Lead: Josh is a software engineer by trade and ML researcher by heart looking to solve the biggest societal problems through education at https://cognition.topace

​We coordinate through slack

​​Discussion doc for this session n/a

​​Docs for previous sessions