Efficient Retrieval with Learned Similarities, Balancing Performance and Cost in LLMs, and More!

Vol.62 for Jul 22 - Jul 28, 2024

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

This week’s newsletter highlights the following research:

1. Optimizing Performance and Cost in Long-Context LLM Applications, from Google DeepMind

2. Bridging the Gap in Long-Context Understanding and RAG Between Open-Access and Proprietary LLMs, from NVIDIA

3. Beyond Dot Products: Efficient Retrieval with Learned Similarities, from Ding et al.

4. An Unsupervised Approach to Measuring Retrieval Complexity in Question Answering, from Amazon

5. A Competitive Mixture-of-Experts Approach to Information Retrieval, from UCAS

6. Leveraging Web Results and GPT-4 for Improved Sponsored Search, from Microsoft

7. Ultra-Long User Behavior Modeling in Recommendation Systems, from Kuaishou

8. Personalizing LLM Prompts for Enhanced Recommendations, from USTC

9. Enhancing Digital Health Chatbots with Query-Based Retrieval Augmented Generation, from Verily

10. Mitigating Polarization in Dynamic Recommender Ecosystems, from TikTok