New Information Retrieval (cs.IR) papers, Monday 19 October 2026
11 papers announced on arXiv. Get them ranked for you, free →
Star Light, Star Bright: Can Multimodal Models Parse an Astronomy Visual Question Answering Dataset Right?
AI summaryThis paper introduces a new astronomy visual question answering dataset containing over two thousand figures and 47,000 questions to evaluate large multimodal models. Testing reveals that current low-cost proprietary models frequently fail at these domain-specific visual reasoning tasks.
AuditKIT: A Unified Evaluation Spine for Benchmarking, Judging, and Comparing Large Language Models and Agentic Systems
AI summaryAuditKIT is an evaluation library that standardizes benchmarking, judging, and tracing for large language models and agents into a single result representation. It enables robust comparisons across diverse workflows by providing unified metrics for tool use, retrieval, and runtime performance.
NoveltyScope: A Multi-Agent Research Assistant for Evidence-Grounded Novelty Assessment of Scientific Papers
AI summaryNoveltyScope is a multi-agent system that assesses the novelty of scientific papers by retrieving prior work, gathering evidence through adversarial LLM roles, and producing a structured report with a novelty score. It achieves lower scoring errors than single-LLM baselines and standard vector similarity on a benchmark of scientific papers.
Improving OCR Accuracy Through Multi-Classifier Verification of Visually Similar Characters
AI summaryThe paper introduces a Character Verification Model that uses binary classifiers for visually confusable pairs to post‑process OCR outputs. It boosts accuracy by 3.95–14.81 pp across three engines, achieving 87.3 % correction of such errors with only 2.3 % false corrections.
Reasoning Reranking with Semantic IDs as the Sole Item Representation
AI summaryThis paper introduces ReRankID to evaluate reasoning rerankers using only compact Semantic IDs instead of text. Results show that targeted post-training enables SID-only models to match text-based LLMs with much shorter prompts, despite producing less faithful item descriptions.
Code Skill Server: A Large-Scale Search Engine for Code Generation Skills
AI summaryCode Skill Server (CSS) is a large-scale search engine that automatically indexes and retrieves reusable code generation skills from many public sources, achieving up to a 13.91% Pass@1 boost and at least 4.21% improvement across diverse models.
Jev-as-a-Judge for Brazilian Portuguese Information Retrieval: Relevance Assessment and Reranking
This study investigates Jev, a System One model designed for fast, structured decision-making, in information retrieval. We use Jev-as-a-judge to generate relevance judgments for four Brazilian Portuguese IR collections covering different domains and compare them with the…
QD-Rec: Open-Ended Evolution of Recommendations via Quality-Diversity
Open-ended search combined with foundation models has produced advances such as faster algorithms, new results on open mathematical problems, and self-improving agents. We extend this paradigm to recommendation for decision-making, focusing on decisions where the set of possible…
LSNR: Learned Sparse News Recommendations with Attentive User Modeling
Current learned sparse approaches for recommender systems, such as STAR, represent a user by concatenating their interaction history into a single query string. For news, this creates two bottlenecks: concatenation collapses a user's distinct reading interests into one flat…
Humanity's Last Choice: A Common Evaluation Standard for Generative Engine Optimization
Humanity's Last Choice describes the idea that the last choice people need to make is which AI system they entrust with the decisions that follow. These systems rely on external sources that content providers can edit through generative engine optimization (GEO) to increase…
Tool Recommendation for Multi-Step Agents: Finding Every Tool a Task Needs
AI summaryTool Recommendation supplies agents with all needed tools via three fixed calls, covering 93% of required tools in Codex sessions. It improves recall over BM25 by 57%, cuts input tokens by 17–40%, and reduces cost by up to 40% without affecting completion.