GPU-Accelerated Feature Interaction for Large-Scale Ad Retrieval, Efficient Neural Sparse Retrieval in Large-Scale Search Engines, and More!

Vol.133 for Dec 01 - Dec 07, 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 Production-Scale LLM Framework for E-Commerce Search Relevance, from Alibaba

2. A Systematic Survey of Deep Research Paradigms in Large Language Models, from Shi et al.

3. Leveraging Semantic Hierarchies in Vector Quantization for Generative Recommendation, from Meta

4. GPU-Accelerated Feature Interaction for Large-Scale Ad Retrieval, from Tencent

5. Balancing Effectiveness and Efficiency in Billion-Scale Sparse Retrieval Models, from Naver

6. A Holistic Framework for Characterizing Text Embedding Spaces, from Rottach et al.

7. Explaining Factual Recall Breakdown in Vision-Language Models, from Venhoff et al.

8. Modeling User Comparison Behavior in Search Ranking Systems, from Airbnb

9. Enhancing Recommendation Systems with LLM-Based Explainable Re-Ranking, from CMU

10. Structure-Aware Retrieval-Augmented Generation for Hierarchical Documents, from CUHK