We Need Improved Data Curation and Attribution in AI for Scientific Discovery
On the map: LLM safety and security → Watermarking generated content and models
Abstract (arXiv)
As the interplay between human-generated and synthetic data evolves, new challenges arise in scientific discovery concerning the integrity of the data and the stability of the models. In this work, we examine the role of synthetic data as opposed to that of real experimental data for scientific research. Our analyses indicate that nearly three-quarters of experimental datasets available on open-access platforms have relatively low adoption rates, opening new opportunities to enhance their discoverability and usability by automated methods. Additionally, we observe an increasing difficulty in distinguishing synthetic from real experimental data. We propose supplementing ongoing efforts in automating synthetic data detection by increasing the focus on watermarking real experimental data, thereby strengthening data traceability and integrity. Our estimates suggest that watermarking even less than half of the real world data generated annually could help sustain model robustness, while promoting a balanced integration of synthetic and human-generated content.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Watermarking generated content and models, one of 12 directions in LLM safety and security. Builds on 30 of its 58 references that are in the map. First posted on arXiv 3 Apr 2025.
Builds on
30 of 58 references.
Closest ancestors · 3
Same direction, newest first
- 2024SoK: Watermarking for AI-Generated Content27 Nov
- 2023A Watermark for Large Language Models24 Jan
- 2022The Ethical Need for Watermarks in Machine-Generated Language7 Sep
Other references · 27
- 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
- 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,243
- 2021Ego4D: Around the World in 3,000 Hours of Egocentric Videocited by 1,530
- 2023InternVid: A Large-scale Video-Text Dataset for Multimodal Understanding and Generationcited by 488
- 2024Scaling Synthetic Data Creation with 1,000,000,000 Personascited by 397
- 2018Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generationcited by 363
- 2023RADAR: Robust AI-Text Detection via Adversarial Learningcited by 187
- 20234M: Massively Multimodal Masked Modelingcited by 118
- 2021Reduced, Reused and Recycled: The Life of a Dataset in Machine Learning Researchcited by 100
- 2023The science of fake newscited by 94
- 2023Unifying Molecular and Textual Representations via Multi-task Language Modellingcited by 92
- 2023Ghostbuster: Detecting Text Ghostwritten by Large Language Modelscited by 90
15 more
- 2023Is GPT-4 a Good Data Analyst?cited by 69
- 2024Strong Model Collapsecited by 62
- 2024Collapse or Thrive? Perils and Promises of Synthetic Data in a Self-Generating Worldcited by 49
- 2024LAB: Large-Scale Alignment for ChatBotscited by 39
- 2024Unraveling Molecular Structure: A Multimodal Spectroscopic Dataset for Chemistrycited by 31
- 2024Navigating Dataset Documentations in AI: A Large-Scale Analysis of Dataset Cards on Hugging Facecited by 27
- 2024Position: Measure Dataset Diversity, Don't Just Claim Itcited by 26
- 2024Universality of the π2/6 Pathway in Avoiding Model Collapsecited by 24
- 2022Accelerating Material Design with the Generative Toolkit for Scientific Discoverycited by 15
- 2023Agent-based Learning of Materials Datasets from Scientific Literaturecited by 15
- 2024Delving into the Utilisation of ChatGPT in Scientific Publications in Astronomycited by 7
- 2024A Little Human Data Goes A Long Waycited by 6
- 2023Shifting Climates: Climate Change Communication from YouTube to TikTokcited by 3
- 2023Domain-agnostic and Multi-level Evaluation of Generative Modelscited by 1
- 2025Making Sense of Data in the Wild: Data Analysis Automation at Scalecited by 1
Built on by
Citing papers per year: 2025: 1.
1 paper.