RIFT: A Scalable Methodology for LLM Accelerator Fault Assessment using Reinforcement Learning.

Khurram Khalil, Muhammad Mahad Khaliq, Khaza Anuarul Hoque

Browse the full DATE paper archive.