A Comprehensive Study of Long Context vs. RAG Performance, Search-o1: Agentic Search-Enhanced Large Reasoning Models, and More!

Vol.86 for Jan 06 - Jan 12, 2025

Stay Ahead of the Curve with the Latest Advancements and Discoveries in Information Retrieval.

This week’s newsletter highlights the following research:

1. A Comprehensive Study of Long Context and RAG Performance, from Li et al.

2. Search-o1: Agentic Search-Enhanced Large Reasoning Models, from Renmin University

3. Scalable Multi-task Adaptation of Retrievers for Enterprise Tasks, from ServiceNow

4. Accelerating Retrieval-Augmented Generation through Pipeline Parallelism, from Amazon

5. A Semantic Uncertainty Approach to Adaptive Knowledge Integration, from Korea University

6. A Systematic Review of Cold-Start Recommendation Systems in the LLM Era, from Zhang et al.

7. Efficient Click-Through Rate Prediction with Graph-Enhanced Retrieval and Adaptive Early Exit LLMs, from Meta

8. A Comprehensive Survey of The Evolution of Vision Language Models from 2019-2024, from UMD

9. A Resource-Efficient Framework for Personalizing Large Language Models, from Google DeepMind

10. Fine-Grained Document Retrieval through Generation-Augmented Learning, from Microsoft