Time (typically): Wed, 2:15PM—3:15PM

Location (typically): WVH366

Group leader: Bryant Curto

Topics: LLM

Untitled

Candidate papers:

• Training

  • Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism (arXiv’20)
  • nnScaler: Constraint-Guided Parallelization Plan Generation for Deep Learning Training (OSDI’24) ⭐
  • MegaScale: Scaling Large Language Model Training to More Than 10,000 GPUs (NSDI’24) ⭐

• Inference

  • ORCA: A Distributed Serving System for Transformer-Based Generative Models (OSDI’22) ⭐
  • PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU (SOSP’24) ⭐
  • SpecInfer: Accelerating Generative Large Language Model Serving with Tree-based Speculative Inference and Verification (ASPLOS’24) ⭐

• Quantization, Low-Rank, Sparsity

  • FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning (ICLR’24)
  • SparseGPT: Massive Language Models Can be Accurately Pruned in One-Shot (ICML’23) ⭐
  • T-MAC: CPU Renaissance via Table Lookup for Low-Bit LLM Deployment on Edge (EuroSys’25) ⭐

• ChatDBG: An AI-Powered Debugging Assistant (arXiv’24) ⭐

• Activation Steering for Robust Type Prediction in CodeLLMs (arXiv’24)

  • authored by Francesca Lucchetti, a PhD student at NU

• Enhancing Code Translation in Language Models with Few-Shot Learning via Retrieval-Augmented Generation (HPEC’24)

• Mapping the Mind of a Large Language Model ⭐

• Automated Unit Test Improvement using Large Language Models at Meta (arXiv'24)

• An Empirical Evaluation of Using Large Language Models for Automated Unit Test Generation (arXiv’23)

  • co-authored by Frank Tip, a professor at NU