RATIO (Retrieval Across Typed Ideation Operations) is a large-scale benchmark designed to evaluate how well retrieval systems can support scientific inspiration. It defines relevance through three ideation moves—Address, Broaden, and Specify—each targeting different levels of abstraction in literature retrieval. The benchmark is built from millions of full-text CS papers using a novel discourse-marker distant supervision method, and includes extensive LLM and human vetting to ensure quality.
By Maayan Sharon, Tom Hope
The paper investigates whether domain-specific fine‑tuning benefits open‑ended scientific reasoning in astronomy. Using a curated 300‑question QA benchmark from 2017–2026 Olympiad‑style materials, the authors compare open‑weight, API‑served general‑purpose, multimodal, and astronomy‑specialized language models. Results show that strong general‑purpose models set the highest correctness baseline, but variations in metric agreement, judge sensitivity, benchmark composition, and modality suggest that domain specialization is task‑ and deployment‑dependent and that domain‑specific evaluation is crucial for scientific workflows.
By Vanessa Lama, Sanjay Das, Emily Herron, Yuan-Sen Ting, Tijmen de Haan, Junqi Yin, Tirthankar Ghosal, Feiyi Wang
The paper introduces the Scientific Contribution Graph, a large-scale resource that extracts 6 million scientific contributions from 655 k open-access papers across multiple disciplines and links them with 36 million prerequisite edges. It frames automated technological roadmapping as the task of identifying contributions and their prerequisites, and presents a new scientific prerequisite prediction task where models forecast which existing technologies enable future discoveries. The authors report that current models achieve a 0.48 MAP score on temporally-filtered backtesting, indicating rapid progress in this area.
By Peter A. Jansen
arXiv:2509. 21028v4 Announce Type: replace Abstract: We introduce SciTrek, a synthetic question-answering dataset for assessing and improving long-context numerical reasoning in large language models (LLMs).
By Miao Li, Alexander Gurung, Irina Saparina, Mirella Lapata
arXiv:2607. 20926v1 Announce Type: new Abstract: Scientific research involves complex information-seeking and reasoning workflows across heterogeneous sources.
By Yinhao Tang, Youqing Fang, Yanan Sun, Wenran Liu, Weiming Zhang, Bin Liu, Kuikun Liu, Wenwei Zhang, Kai Chen
SHELF is a Python system that creates synthetic, controlled benchmark data for evaluating large language models on bibliographic tasks such as classification, clustering, retrieval, pair classification, and instruction retrieval. It generates 62,899 model-written documents based on Library of Congress vocabularies and compares methods like TF, TF‑IDF, BM25, popular encoders, and zero‑shot decoders, reporting performance metrics such as 0.8887 for subject classification and 0.2605 for genre‑form classification. The tool also allows independent variation of bibliographic facets and can produce unseen documents beyond a model’s training cutoff, with results indicating that model rankings transfer more reliably than absolute scores when compared to other benchmarks.
By Michael J. Bommarito II