/Outstanding-Paper

[Paper] This repositories review outstanding papers in various topics

Outstanding Papers

LLM and RAG

Enhancing Q&A Text Retrieval with Ranking Models: Benchmarking, fine-tuning and deploying Rerankers for RAG (NVIDIA-AI, 2024)

  • The paper introduces a state-of-the-art ranking model, NV-RerankQAMistral-4B-v3, which achieves a significant accuracy increase of 14% compared to pipelines with other rerankers.

When “A Helpful Assistant” Is Not Really Helpful: Personas in System Prompts Do Not Improve Performances of Large Language Models (Carnegie Mellon University, 2024)

  • The paper presents a systematic evaluation of personas in system prompts.
  • The findings demonstrates that while adding a persona may lead to performance gains in certain settings, the effect of each persona can be largely random.

LLMs + Persona-Plug = Personalized LLMs (Renmin University of China, 2024)

Medical Imaging

DenoMamba: A fused state-space model for low-dose CT denoising (Bilkent University, 2024)
Code

  • The paper proposes DenoMamba, a novel deep learning model for low-dose CT denoising.
  • Among the earliest state-space models (SSM) for image denoising, DenoMamba fuses spatial and channel SSM modules within a unified hourglass architecture to achieve high-quality reconstruction of CT images from low-dose measurements.