Large language models (LLMs) are increasingly used to generate research ideas from scientific literature, but the impact of the retrieval strategy on the ideas produced is unclear. This study isolates that contribution by comparing five retrieval strategies and a no‑literature baseline across 45,000 ideas in 15 machine learning topics and ten open‑weight models. The results show that while all retrieval strategies reduce diversity within a single set of papers, they increase diversity when ideas from different sets are pooled; effects on quality and novelty are modest but topic‑dependent, and a JSD‑based attribution method reveals that most ideas are driven by one or two key papers, with more varied source attribution correlating with greater idea diversity.
arXiv:2607. 01233v1 Announce Type: cross Abstract: LLMs are increasingly used to brainstorm research ideas, but existing evaluations mostly judge individual ideas by novelty, feasibility, or expert preference.
arXiv:2610.17439v1 Announce Type: new Abstract: The success of autonomous research systems relies on the quality of the scientific hypotheses they generate. Recent studies report that human reviewers...
arXiv:2606. 26130v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used to guide research methodology, yet their default methodological tendencies under minimal prompting remain unclear.
The paper introduces a method for training large language models to generate diverse, high‑quality scientific ideas grounded in input documents. By using a set‑level reinforcement learning objective that rewards ideas for being distinct from earlier ones, the authors train an 8B model that outperforms Claude Opus 4.8 and GPT‑6 Astra on a contamination‑free benchmark. Human experts rate the model’s idea sets as equally promising while containing slightly more unique ideas, and automated experiments show comparable executability.
arXiv:2603.01092v3 Announce Type: replace-cross Abstract: Scientific discovery is constrained not only by what is true, but by what is cognitively available to researchers. Many directions are cohere...
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.
arXiv:2606. 12071v1 Announce Type: cross Abstract: LLMs are increasingly used to generate and judge scientific ideas.
The paper "Learning to Ideate for Scientific Impact" explores using delayed signals of scientific uptake—specifically citation-normalized impact—as feedback to steer large language models toward generating high‑impact research ideas. The authors build a dataset of over 100,000 computer science papers, train a reward model to predict citation impact from goal‑idea pairs, and align an idea generator via supervised fine‑tuning and reinforcement learning. Evaluation with a reference‑grounded protocol shows that the RL‑tuned model consistently produces ideas with higher estimated impact than baseline models.
arXiv:2606. 25198v2 Announce Type: replace Abstract: Autonomous AI Research promises to accelerate the scientific progress of machine learning.
SciRerankBench is a benchmark designed to evaluate rerankers in retrieval‑augmented large language models (RAG‑LLMs) for scientific literature question answering. It covers five scientific subjects and introduces three types of question‑context‑answer pairs—Noisy Contexts, Semantically Similar but Logically Irrelevant Contexts, and Counterfactual Contexts—to test noise resilience, relevance disambiguation, and factual consistency. The study evaluates 13 popular rerankers across five LLM families, offering detailed insights into their strengths and limitations.
The paper investigates whether large language models (LLMs) can reliably assess scientific hypotheses by using a logit-based energy scoring method that leverages the model’s intrinsic confidence. Across 1,323 papers in 12 disciplines, this intrinsic scoring achieved a 33.0% Hit@1 rate, outperforming a prompted listwise ranking approach that scored 16.6%. The best result, a 1‑billion‑parameter model with logit-based energy scoring, reached 53.1% Hit@1, suggesting that confidence‑based evaluation could improve trustworthy AI‑enabled scientific discovery.
arXiv:2610.02022v2 Announce Type: replace Abstract: AI-based ideation systems are often evaluated on the novelty of their ideas, and that judgment is increasingly delegated to large language models....