Evaluating Multilingual Metadata Quality in Crossref

Crossref: The sustainable source of community-owned scholarly metadata

This paper describes the scholarly metadata collected and made available by Crossref, as well as its importance in the scholarly research ecosystem. Containing over 106 million records and expanding at an average rate of 11% a year, Crossref’s metadata has become one of the major sources of scholarly data for publishers, authors, librarians, funders, and researchers. The metadata set consists of 13 content types, including not only traditional types, such as journals and conference papers, but also data sets, reports, preprints, peer reviews, and grants. The metadata is not limited to basic publication metadata, but can also include abstracts and links to full text, funding and license information, citation links, and the information about corrections, updates, retractions, etc. This scale and breadth make Crossref a valuable source for research in scientometrics, including measuring the growth and impact of science and understanding new trends in scholarly communications. The metadata is available through a number of APIs, including REST API and OAI-PMH. In this paper, we describe the kind of metadata that Crossref provides and how it is collected and curated. We also look at Crossref’s role in the research ecosystem and trends in metadata curation over the years, including the evolution of its citation data provision. We summarize the research used in Crossref’s metadata and describe plans that will improve metadata quality and retrieval in the future.

Manuscript submission systems and metadata completeness in Crossref: Patterns and associations

The importance of open research information, particularly publication metadata, is widely recognised. Crossref is one of the most important infrastructures for registering open metadata as part of DOI record registration. It is widely known, however, that the metadata of many publications is far from complete, with many publishers making certain metadata openly available, but failing to do so for other metadata elements. Publishers’ ability to register this metadata with Crossref depends on their capacity to capture and retain this data in their production workflows. Manuscript submission systems are an important, yet largely overlooked, factor in the extent to which publishers make metadata available through Crossref. In this paper, we present the results of an analysis investigating the relation between the level of metadata that publishers deposit with Crossref and the submission systems that they deploy for their journals. We have looked at the 153 publishers with the largest amounts of publications in Crossref and concentrate on the four most commonly used systems: Editorial Manager, ScholarOne, Open Journal Systems (OJS) and eJournalPress. We show that some submission systems appear better suited to capturing certain metadata elements. However, there are always cases where publishers using the same system differ widely in the level of metadata they register, suggesting that technology is not the only prohibiting factor and other considerations are at play.

Wiki-40B: Multilingual Language Model Dataset

We propose a new multilingual language model benchmark that is composed of 40+ languages spanning several scripts and linguistic families. With around 40 billion characters, we hope this new resource will accelerate the research of multilingual modeling. We train monolingual causal language models using a state-of-the-art model (Transformer-XL) establishing baselines for many languages. We also introduce the task of multilingual causal language modeling where we train our model on the combined text of 40+ languages from Wikipedia with different vocabulary sizes and evaluate on the languages individually. We released the cleaned-up text of 40+ Wikipedia language editions, the corresponding trained monolingual language models, and several multilingual language models with different fixed vocabulary sizes.

False Friends Are Not Foes: Investigating Vocabulary Overlap in Multilingual Language Models

Subword tokenizers trained on multilingual corpora naturally produce overlapping tokens across languages. Does token overlap facilitate cross-lingual transfer or instead introduce interference between languages? Prior work offers mixed evidence, partly due to varied setups and confounders, such as token frequency or subword segmentation granularity. To address this question, we devise a controlled experiment where we train bilingual autoregressive models on multiple language pairs under systematically varied vocabulary overlap settings. Crucially, we explore a new dimension to understanding how overlap affects transfer: the semantic similarity of tokens shared across languages. We first analyze our models' hidden representations and find that overlap of any kind creates embedding spaces that capture cross-lingual semantic relationships, while this effect is much weaker in models with disjoint vocabularies. On XNLI and XQuAD, we find that models with overlap outperform models with disjoint vocabularies, and that transfer performance generally improves as overlap increases. Overall, our findings highlight the advantages of token overlap in multilingual models and show that substantial shared vocabulary remains a beneficial design choice for multilingual tokenizers.

When Trivia Is Not Trivial: Everyday Knowledge Failures in Multilingual LLMs

Quiz rooms, trivia nights, and quiz shows challenge human knowledge across a wide range of topics, from canonical facts to everyday culture. In this paper, we examine whether large language models (LLMs) can perform competitively in such settings, using quiz-style questions to test them on both common and niche topics. We introduce TriviaRoomQA, a multilingual benchmark designed to evaluate everyday, culturally grounded, and long-tail knowledge across 288 topics. The benchmark contains 3,300 parallel multiple-choice questions in six European languages and additional 5,340 French-only questions for a more fine-grained case study. We evaluate 30 open-weight LLMs from European, Asian, and North American providers, covering models from 7 to 70B parameters. We find that models are strong on knowledge-intensive topics such as history, geography, and mathematics, but substantially weaker on everyday popular-culture topics such as celebrities, music, movies, and news. Moreover, model performance varies across languages even for the same underlying questions, suggesting that access to factual knowledge is not always language-independent. In sum, our dataset and experiments demonstrate an important knowledge gap which is not captured by existing academic-based saturated benchmarks.

Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpus

Large language models have led to remarkable progress on many NLP tasks, and researchers are turning to ever-larger text corpora to train them. Some of the largest corpora available are made by scraping significant portions of the internet, and are frequently introduced with only minimal documentation. In this work we provide some of the first documentation for the Colossal Clean Crawled Corpus (C4; Raffel et al., 2020), a dataset created by applying a set of filters to a single snapshot of Common Crawl. We begin by investigating where the data came from, and find a significant amount of text from unexpected sources like patents and US military websites. Then we explore the content of the text itself, and find machine-generated text (e.g., from machine translation systems) and evaluation examples from other benchmark NLP datasets. To understand the impact of the filters applied to create this dataset, we evaluate the text that was removed, and show that blocklist filtering disproportionately removes text from and about minority individuals. Finally, we conclude with some recommendations for how to created and document web-scale datasets from a scrape of the internet.

Scientific production in the era of Large Language Models

Large Language Models (LLMs) are rapidly reshaping scientific research. We analyze these changes in multiple, large-scale datasets with 2.1M preprints, 28K peer review reports, and 246M online accesses to scientific documents. We find: 1) scientists adopting LLMs to draft manuscripts demonstrate a large increase in paper production, ranging from 23.7-89.3% depending on scientific field and author background, 2) LLM use has reversed the relationship between writing complexity and paper quality, leading to an influx of manuscripts that are linguistically complex but substantively underwhelming, and 3) LLM adopters access and cite more diverse prior work, including books and younger, less-cited documents. These findings highlight a stunning shift in scientific production that will likely require a change in how journals, funding agencies, and tenure committees evaluate scientific works.

Cross-Lingual Exploration for Parametric Knowledge | Idan Szpektor

Accepted to EMNLP Findings! While Large Language Models encode vast amounts of multicultural and factual information, this parametric knowledge is often unevenly accessible across languages. Standard inference frequently fails to surface localized facts, creating persistent gaps in cross-lingual knowledge transfer and consistency. We address this challenge in our paper "Cross-Lingual Exploration for Parametric Knowledge" (https://lnkd.in/drf36bV7), a collaboration between Elisha Diskind, Itamar Trainin, and Omri Abend from The Hebrew University of Jerusalem, Leshem Choshen from the Weizmann Institute of Science, alongside Uri Shaham and myself from our Google Research IL group. We formalize cross-lingual exploration as a structured search process across four core dimensions: language selection, exploration routing, answer aggregation, and inference budget. Evaluating across 17 typologically diverse languages on the ECLeKTic and CLIKE benchmarks, we show that allowing models to autonomously navigate alternative linguistic paths yields up to a 21% gain in knowledge transfer and a 16% boost in factual recall over native baselines, surpassing both standard English-pivot routing and native-language reasoning. Crucially, cross-lingual exploration defines a significantly more efficient compute Pareto frontier than scaling within the native query language, while driving intrinsic cross-lingual consistency gains beyond what accuracy improvements alone explain. Technically, this moves the needle for multilingual inference and knowledge elicitation pipelines, demonstrating that strategic language switching is a powerful, training-free mechanism for unlocking latent parametric knowledge.

The diffusion of large language models in published academic articles

Large language models (LLMs) are rapidly changing academic research, raising questions of who is adopting these tools and under what conditions. This article analyzes full texts of 7.3 million journal articles published from 2020–2025 by four major publishers (Elsevier, Frontiers, MDPI, and PLoS) to track the prevalence of LLM-associated language and identify social and institutional correlates of adoption. A corpus of 228 focal words exhibiting sharp post-2022 frequency increases consistent with LLM output was developed; articles were scored on their rate of focal word usage. By 2025, an estimated 57% of published articles exhibited evidence of LLM influence, up from 12% in 2023. Among articles exhibiting LLM-influenced text, there is substantial heterogeneity, ranging from subtle linguistic influence to articles mostly or entirely LLM-generated. Difference-in-differences models reveal that LLM-associated language varies markedly across regions, institutional ranks, publishers, disciplines, and journal tiers. Economic development and proximity to English as a primary language are key predictors of regional variation. Lower-ranked institutions exhibit higher rates than elite universities, young for-profit publishers show elevated rates vis-à-vis competitors, and academic fields differ widely in adoption. LLM adoption in academic writing is pervasive but socially stratified. As models grow more powerful and their use becomes further entrenched in academic research, understanding social dynamics of adoption will be essential for governing the evolving relationship between AI and academic knowledge production.

Two billion citation links in Crossref help research travel further - Crossref

We’ve recently reached an important milestone for the research nexus: the works in our metadata corpus are now connected with over 2 billion citation links! This is a great opportunity to share a dedicated dataset and discuss why these are important for science.

You are Crossref - Crossref

Crossref runs open infrastructure to link research objects, entities, and actions—creating a lasting and reusable scholarly record that underpins open science. Together with our >25,000 members in 167 countries, we drive metadata exchange and support 2.1 billion monthly API queries, facilitating global research communication, for the benefit of society.

Post-Training Language Models for Crosslingual Consistency

Language models often respond inconsistently to translation-equivalent prompts across languages, undermining the reliability of multilingual systems. To quantify this, we give an information-theoretic definition of crosslingual consistency as a divergence bound between a model's response distribution and its round-trip pushforward across languages. We then introduce penalized consistency optimization (PCO), a post-training procedure that couples this divergence with a Kullback-Leibler penalty to a fixed reference language model. Because direct optimization of PCO requires expensive on-policy roll-outs, we propose a tractable surrogate, direct consistency optimization (DCO), which can be optimized off-policy. Across diverse language models and 26 languages, DCO significantly improves crosslingual consistency, outperforms existing methods, and enables targeted alignment of low-resource languages.

Incorrect Citation Association for Articles in Online-Only Springer Nature Journals

We show that citation metrics of journal articles in many of the online-only Springer Nature journals and associated ones are distorted, going back to articles from 2001. We find that most likely due to an API response error, there are many incorrect references which typically lead to Article Number 1 of a given Volume. Among others, the issue affects journals such as Scientific Reports, Nature Communications, Communications journals, Cell Death & Disease, Light: Science & Applications, as well as many BMC, Discovery and npj journals. Beyond the negative effect of introducing incorrect reference information, this distorts the citation statistics of articles in these journals, with a few articles being massively over-cited compared to their peers, while many lose citations; e.g. both in Scientific Reports and in Nature Communications, 5 of the 10 top cited articles have article numbers of 1. We validate the distorted statistics by assessing data from multiple scientific literature databases: Crossref, OpenCitations, Semantic Scholar, and the journals' websites. The issue primarily arises from the inconsistent transition from page-based referencing of articles to article number-based referencing, as well as the improper handling of the change in the publisher's article metadata API. It seems that the most pressing problem has been present since approximately 2011, which we estimate affects the citation count of millions of authors.

DBpedia Abstracts: A Large-Scale, Open, Multilingual NLP Training Corpus

The ever increasing importance of machine learning in Natural Language Processing is accompanied by an equally increasing need in large-scale training and evaluation corpora. Due to its size, its openness and relative quality, the Wikipedia has already been a source of such data, but on a limited scale. This paper introduces the DBpedia Abstract Corpus, a large-scale, open corpus of annotated Wikipedia texts in six languages, featuring over 11 million texts and over 97 million entity links. The properties of the Wikipedia texts are being described, as well as the corpus creation process, its format and interesting use-cases, like Named Entity Linking training and evaluation.

A large-scale audit of dataset licensing and attribution in AI

The race to train language models on vast, diverse and inconsistently documented datasets raises pressing legal and ethical concerns. To improve data transparency and understanding, we convene a multi-disciplinary effort between legal and machine learning experts to systematically audit and trace more than 1,800 text datasets. We develop tools and standards to trace the lineage of these datasets, including their source, creators, licences and subsequent use. Our landscape analysis highlights sharp divides in the composition and focus of data licenced for commercial use. Important categories including low-resource languages, creative tasks and new synthetic data all tend to be restrictively licenced. We observe frequent miscategorization of licences on popular dataset hosting sites, with licence omission rates of more than 70% and error rates of more than 50%. This highlights a crisis in misattribution and informed use of popular datasets driving many recent breakthroughs. Our analysis of data sources also explains the application of copyright law and fair use to finetuning data. As a contribution to continuing improvements in dataset transparency and responsible use, we release our audit, with an interactive user interface, the Data Provenance Explorer, to enable practitioners to trace and filter on data provenance for the most popular finetuning data collections: www.dataprovenance.org.

Language agents achieve superhuman synthesis of scientific knowledge

Language models are known to hallucinate incorrect information, and it is unclear if they are sufficiently accurate and reliable for use in scientific research. We developed a rigorous human-AI comparison methodology to evaluate language model agents on real-world literature search tasks covering information retrieval, summarization, and contradiction detection tasks. We show that PaperQA2, a frontier language model agent optimized for improved factuality, matches or exceeds subject matter expert performance on three realistic literature research tasks without any restrictions on humans (i.e., full access to internet, search tools, and time). PaperQA2 writes cited, Wikipedia-style summaries of scientific topics that are significantly more accurate than existing, human-written Wikipedia articles. We also introduce a hard benchmark for scientific literature research called LitQA2 that guided design of PaperQA2, leading to it exceeding human performance. Finally, we apply PaperQA2 to identify contradictions within the scientific literature, an important scientific task that is challenging for humans. PaperQA2 identifies 2.34 +/- 1.99 contradictions per paper in a random subset of biology papers, of which 70% are validated by human experts. These results demonstrate that language model agents are now capable of exceeding domain experts across meaningful tasks on scientific literature.