Perturbation analysis of Large Language Models (LLM) [Microsoft]
An episode of the Snacks Weekly on Data Science podcast, hosted by Pan Wu, titled "Perturbation analysis of Large Language Models (LLM) [Microsoft]" was published on March 25, 2024 and runs 14 minutes.
March 25, 2024 ·14m · Snacks Weekly on Data Science
Summary
Description: In this episode, we discuss how the data science team at Microsoft designed perturbation analysis to understand the Large Language Model’s performance on commonly seen tasks. Based on their published tech blog, with the link provided here for your reference: https://medium.com/data-science-at-microsoft/perturbation-analysis-and-llms-how-sensitive-are-llms-to-their-input-91a8407a971f
Episode Description
Description: In this episode, we discuss how the data science team at Microsoft designed perturbation analysis to understand the Large Language Model’s performance on commonly seen tasks.
Based on their published tech blog, with the link provided here for your reference:
Similar Episodes
Jun 19, 2025 ·46m
Jun 13, 2025 ·40m
May 20, 2025 ·80m
May 13, 2025 ·74m
May 7, 2025 ·64m