Storage-Efficient Multi-Vector Retrieval, Enhancing Negation Detection in Text Embeddings, and More!

Vol.98 for Mar 31 - Apr 06, 2025

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

This week’s newsletter highlights the following research:

1. Enhancing Sequential Recommendation with Multi-Step Inference-Time Reasoning, from Tang et al.

2. Bridging Large Language Models and Recommendation Systems through Reinforcement Learning, from Lin et al.

3. Test-Time Parameter Injection for Efficient Knowledge Fusion in LLMs, from Tan et al.

4. Optimizing Passage Ranking Efficiency Through Small-Large Model Collaboration, from Baidu

5. Enhancing Negation Awareness in Universal Text Embeddings, from Cao et al.

6. Efficient Fixed-Vector Representation for Multi-Vector Retrieval, from MacAvaney et al.

7. Enhancing LLMs with Insight-Driven Knowledge Augmentation, from Megagon Labs

8. Bridging Queries and Products with Aligned Text Identifiers for E-commerce Retrieval, from JD[.com]

9. Enhancing Linear Recurrent Models with Context-Aware Retrieval, from Wang et al.

10. Enhancing Retrieval-Augmented Language Models with Shared Context Attribution, from Xu et al.