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

This week’s newsletter highlights the following research:

1. Measuring the Efficiency Gap Between Human and Agent Document Reasoning, from Borchmann et al.

2. Improving Search Agent with One Line of Code, from Tencent

3. BM25-V: Sparse Auto-Encoder Visual Word Scoring for Image Retrieval, from Han et al.

4. Building Uber Eats’ Multilingual Semantic Retrieval Pipeline, from Uber

5. Generative Embeddings from Large Language Models, from BehnamGhader et al.

6. Differentiable Isotonic Regression as a Plug-and-Play Calibration Layer for Deep Recommendation Models, from LinkedIn

7. Linking Retrieval Metrics to Generation Quality in RAG Systems, from Samuel et al.

8. Agent-Driven Deep Research as the Next Paradigm for Recommender Systems, from Ou et al.

9. Hierarchical Agent Coordination for Scalable Multi-Document QA, from Akay et al.

10. Teaching Small Models to Navigate Large Tool Ecosystems, from Microsoft