Daily edition · Information Retrieval
TILens turns technical updates into a focused daily brief: official releases, trusted reporting, and practitioner analysis, deduplicated and organized by topic.
30 articles · 4 sources · 30 papers ·
Top topics: Information Retrieval · AI · AI Research
arXiv:2609.01913v2 Announce Type: replace Abstract: In product entity resolution, relationship definitions constantly evolve with business needs, yet adapting to each change traditionally requires slow, costly human…
arXiv:2510.18239v4 Announce Type: replace Abstract: Scaling large recommendation systems requires advancing three major frontiers: processing longer user histories, expanding candidate sets, and increasing model…
arXiv:2610.12243v1 Announce Type: new Abstract: Dense retrieval usually ranks text chunks by their semantic similarity to a question. This ignores structure that many data systems already store, including section…
arXiv:2610.11598v1 Announce Type: new Abstract: In this paper, we provide an overview of the NTCIR-19 Automatic Evaluation of LLMs 2 (AEOLLM-2) task. Building on the success of the NTCIR-18 core task AEOLLM, we proposed…
arXiv:2610.10556v1 Announce Type: new Abstract: Cascaded recommender systems use the same user history for retrieval and ranking, while the two stages have access to different information at different points of an…
arXiv:2610.11666v1 Announce Type: new Abstract: Universal multimodal retrieval typically encodes a query once and ranks independently indexed items by embedding similarity. This design supports efficient search, but…
arXiv:2610.11650v1 Announce Type: new Abstract: Similar agent skills can share instructions but differ in their conditions of use. Query-based text selection may retain shared instructions and omit these distinctions.…
arXiv:2610.11553v1 Announce Type: new Abstract: Accurate and scalable visual document retrieval (VDR) requires both fine-grained page understanding and efficient indexing, yet existing approaches struggle to achieve…
arXiv:2610.08077v2 Announce Type: replace-cross Abstract: Reinforcement learning with verifiable rewards (RLVR) turns agent experience into learning signals primarily through scalar outcome rewards after interaction.…
arXiv:2610.12256v1 Announce Type: cross Abstract: Omnimodal embeddings naturally involve both shared representations and modality-specific features across heterogeneous inputs. However, existing omnimodal embedding…
arXiv:2603.07233v2 Announce Type: replace-cross Abstract: Predicting transcriptional responses to genetic perturbations is fundamental to functional genomics and therapeutic discovery. Recent deep learning models have…
arXiv:2610.11270v1 Announce Type: cross Abstract: Personalized conversational AI relies on long-term memory systems that extract facts from user utterances and store them in persistent vector stores. Despite progress in…
arXiv:2610.11370v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) grounds language models in external corpora. Agentic RAG enables iterative search, yet exposes the model to isolated chunks without…
arXiv:2610.10920v1 Announce Type: new Abstract: E-commerce search pages are critical touchpoints for millions of online shoppers. While traditional search engines return a ranked list of results, modern E-commerce…
arXiv:2610.11123v1 Announce Type: new Abstract: Most human-agent interaction today remains text-based. Natural language can impose cognitive overload, ambiguity, information chaos, and slow input for complex tasks;…
arXiv:2610.12312v1 Announce Type: new Abstract: Large-scale information retrieval systems, including retrieval-augmented generation (RAG) and recommendation engines, widely use multi-layered hierarchical data structures…
arXiv:2606.10388v4 Announce Type: replace Abstract: A skill can match a task's topic while conflicting with its resource, procedure, or output requirements. We study this as same-capability risk-exposure retrieval and…
arXiv:2610.11922v1 Announce Type: new Abstract: Project Greenhouse represents our exploration of a simple thesis: We believe that it is possible to build fully open and sovereign models for agentic search with only…
arXiv:2610.11506v1 Announce Type: cross Abstract: Graph clustering is a fundamental task in data analysis, aiming at grouping nodes with similar characteristics in the graph into clusters. This problem has been widely…
arXiv:2610.11439v1 Announce Type: cross Abstract: Before Ethereum made the "oracle problem" a household term, Bitcoin already had oracles serving as feeds, key-release services, federated signers, and arbiters that…
Showing 1 day · 30 items available
