arXiv:2604.20851v1 Announce Type: new
Abstract: Modern video-text retrieval (VTR) models excel on in-distribution benchmarks but are highly vulnerable to real-world query shifts, where the distribution of query data deviates from the training domain, leading to a sharp performance drop. Existing image-focused robustness solutions are inadequate to handle this vulnerability in video, as they fail to address the complex spatio-temporal dynamics inherent in these shifts. To systematically evalu...
arXiv:2604.20850v1 Announce Type: new
Abstract: Dense retrieval systems rank passages by embedding similarity to a query, but multi-hop questions require passages that are associatively related through shared reasoning chains. We introduce Association-Augmented Retrieval (AAR), a lightweight transductive reranking method that trains a small MLP (4.2M parameters) to learn associative relationships between passages in embedding space using contrastive learning on co-occurrence annotations. At ...
arXiv:2604.20849v1 Announce Type: new
Abstract: Retrieval-augmented generation over semi-structured sources such as HTML is constrained by a mismatch between document structure and the flat, sequence-based interfaces of today's embedding and generative models. Retrieval pipelines often linearize documents into fixed-size chunks before indexing, which obscures section structure, lists, and tables, and makes it difficult to return small, citation-ready evidence without losing the surrounding c...
arXiv:2604.20848v1 Announce Type: new
Abstract: Large Language Model (LLM)-based recommendation systems have demonstrated remarkable capabilities in understanding user preferences and generating personalized suggestions. However, existing approaches face critical challenges in transparency, knowledge grounding, and the ability to provide coherent explanations that foster user trust. We introduce MATRAG (Multi-Agent Transparent Retrieval-Augmented Generation), a novel framework that combined ...
arXiv:2604.20847v1 Announce Type: new
Abstract: Music Recommendation Systems (MRSs) are a cornerstone of modern streaming platforms. Existing recommendation models, spanning both recall and ranking stages, predominantly rely on collaborative filtering, which fails to exploit the intrinsic characteristics of audio and consequently leads to suboptimal performance, particularly in cold-start scenarios. However, existing music recommendation datasets often lack rich multimodal information, such ...
arXiv:2604.20846v1 Announce Type: new
Abstract: Next point-of-interest (POI) recommendation requires modeling user mobility as a spatiotemporal sequence, where different behavioral factors may evolve at different temporal and spatial scales. Most existing methods compress a user's history into a single latent representation, which tends to entangle heterogeneous signals such as routine mobility patterns, short-term intent, and temporal regularities. This entanglement limits the flexibility o...
arXiv:2604.20844v1 Announce Type: new
Abstract: Recent GraphRAG methods integrate graph structures into text indexing and retrieval, using knowledge graph triples to connect text chunks, thereby improving retrieval coverage and precision. However, we observe that treating text chunks as the basic unit of knowledge representation rigidly groups multiple atomic facts together, limiting the flexibility and adaptability needed to support diverse retrieval scenarios. Additionally, triple-based en...
Most GUCs in this series will be operationally irrelevant to most readers. This one is not. application_name is the single cheapest piece of observability infrastructure PostgreSQL ships, and an astonishing number of production databases are running with it unset or stuck at a client library&rsqu…
A benchmark came out of AWS earlier this month showing PostgreSQL throughput on Linux 7.0 dropping to 0.51x what the same workload produced on Linux 6.x. The Phoronix headline wrote itself. Hacker News did what Hacker News does. By the end of the week, I had been asked by three separate clients w…
Expanding CS education in prison is a vote toward a future where people with experience of incarceration are represented among computing faculty and industry.
Amazon's Brand Registry de-centers brands as a consumer search mechanism and will likely amplify Amazon’s own power by making its search function and algorithm more important in finding products.
Postgres Conference 2026 was held in San Jose, California, and once again, I was lucky to be invited to speak. This is a great show for the 'hallway track' where you talk to members of the community and discover many interesting things.I had a brief conversation with two early contributors to the original PostgreSQL project. One said he was surprised by how much of his code was still in the code base after FORTY YEARS. AI and MCP are fully interlaced in many projects. The one that struck ...
RAG
