Ryan Anthony Marinelli

• Marinelli, Ryan Anthony (2026). Bayesian Benchmarking: Applications of Statistics for Efficient Model Ranking. In Bin, Riccardo De; Salucci, Clara Bertinelli & Hubin, Aliaksandr (Ed.), Proceedings of the 40th International Workshop on Statistical Modelling 28.06 – 03.07.2026, Oslo, Norway. International Workshop on Statistical Modelling. ISSN 9788255325000. doi: https:/www.mn.uio.no/math/english/research/groups/statistics-data-science/events/conferences/iwsm2026/proceedings/proceedings_iwsm2026_proofonline.pdf. Full text in Research Archive Show summary

  This work explores more efficient benchmarking techniques to com- pare the performance between models. By leveraging Bayesian techniques, it is possible to confidently rank models by their performance without excessive compute.

• Marinelli, Ryan Anthony (2025). Adversarial Prompt Filtering Based on Societal Risk, 25th International Symposium on Cluster, Cloud and Internet Computing Workshops (CCGridW). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798331509385. p. 225–228. doi: 10.1109/CCGridW65158.2025.00043. Full text in Research Archive
• Marinelli, Ryan Anthony & Eckhoff, Magnus Wiik (2025). Leaking LoRa: An Evaluation of Password Leaks and Knowledge Storage in Large Language Models. arXiv. doi: 10.48550/arXiv.2504.00031. Full text in Research Archive
• Marinelli, Ryan Anthony; Bisztray, Tamas György & Pichlmeier, Josef (2025). Harnessing chain-of-thought metadata for task routing and adversarial prompt detection. arXiv. doi: 10.48550/arXiv.2503.21464. Full text in Research Archive
• Marinelli, Ryan Anthony (2025). Responsible Development of Offensive AI. arXiv. doi: 10.48550/arXiv.2504.02701. Full text in Research Archive
• Marinelli, Ryan Anthony; Erdodi, Laszlo Tibor; Sommervoll, Åvald Åslaugson & Zennaro, Fabio Massimo (2024). Extending Q-Learning Agents in SQLi Environments. In Ukjent (Eds.), 2024 Cyber Awareness and Research Symposium (CARS). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350386417. doi: 10.1109/cars61786.2024.10778889. Full text in Research Archive
• Norbert, Tihanyi; Bisztray, Tamas György; Dubniczky, Richard; Toth, Rebeka; Borso, Bertalan & Cherif, Bilel [Show all 14 contributors for this article] (2024). Dynamic intelligence assessment: Benchmarking llms on the road to agi with a focus on model confidence. In Ding, Wei (Eds.), 2024 IEEE International Conference on Big Data, Dec 15 - Dec 18, 2024 • Washington DC, USA - Proceedings. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350362480. doi: 10.1109/BigData62323.2024.10825051. Full text in Research Archive
• Marinelli, Ryan Anthony; Sommervoll, Åvald Åslaugson & Erdodi, Laszlo Tibor (2024). Optimizing Deployment of Homomorphic Encryption and SQL using Reinforcement Learning. In Ding, Wei (Eds.), 2024 IEEE International Conference on Big Data, Dec 15 - Dec 18, 2024 • Washington DC, USA - Proceedings. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350362480. doi: 10.1109/BigData62323.2024.10825773. Full text in Research Archive
• Marinelli, Ryan Anthony (2024). Causal Tracing to Identify Hacking Knowledge in Large Language Models. In Ding, Wei (Eds.), 2024 IEEE International Conference on Big Data, Dec 15 - Dec 18, 2024 • Washington DC, USA - Proceedings. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350362480. doi: 10.1109/BigData62323.2024.10825125. Full text in Research Archive
• Marinelli, Ryan Anthony (2023). Henrik Skaug Sætra (ed.) Technology and Sustainable Development: The Promise and Pitfalls of Techno-Solutionism. Prometheus. ISSN 0810-9028. doi: 10.13169/prometheus. Full text in Research Archive

View all works in NVA

• Halnes, Geir; Javaid, Anam; Solvang, Michael; Nichele, Stefano; Riegler, Michael & Singstad, Bjørn-Jostein [Show all 82 contributors for this article] (2025). Proceedings of NORA’s annual conference 2025. Nordic Machine Intelligence (NMI). 5(3), p. 1–14. doi: 10.5617/nmi.12551. Full text in Research Archive
• Marinelli, Ryan Anthony (2025). AI & Law. Full text in Research Archive Show summary

  Lecture slides found at https://docs.google.com/presentation/d/1WjI3O96JuGNTHr-gsTqLeL7ytvrcnOiQPXY6LudINz8/edit?usp=sharing

• Marinelli, Ryan Anthony & Asefaw, Million (2025). Applications of PINNs and GPS Jamming. Full text in Research Archive Show summary

  Slides from presentation may be found at https://docs.google.com/presentation/d/1FIIzh8cS9kYV1FWmvyzX8JBsxDbLjuSpNYBYS1O3_Cc/edit?usp=sharing

• Marinelli, Ryan Anthony (2025). Chess Analysis. Full text in Research Archive Show summary

  Slides can be found at https://docs.google.com/presentation/d/1N6UH6n65juf34qCnpmAVnwHXRLNfqO17-ZRVbbMVFIk/edit?usp=sharing

• Salenko, Yaroslaw; Marinelli, Ryan Anthony & Lammle, Markus (2025). Holographic Imaging in Radiology: Principles and the Role of AI in Image Generation. Full text in Research Archive Show summary

  Poster can be found at https://drive.google.com/file/d/1swNRIT7FGbSWmINEM2Ukbfe5G9nZAO0w/view

• Marinelli, Ryan Anthony (2025). Responsible Development of Offensive AI. Full text in Research Archive Show summary

  Link to presentation https://youtu.be/B_S-3-2Ns4s

• Marinelli, Ryan Anthony (2025). Cybersecurity DND. Full text in Research Archive Show summary

  Event conducted in preparation for Cyber Operations course lecture. The event was advertised via SIFI. Slides located at https://docs.google.com/presentation/d/12QrQkVjVI-LYptpNj0yHYS6vg2T7uVRpYPtui8VvT6Q/edit?usp=sharing

• Marinelli, Ryan Anthony (2025). Tabletop Simulation. Full text in Research Archive Show summary

  Slides may be found at https://docs.google.com/presentation/d/12QrQkVjVI-LYptpNj0yHYS6vg2T7uVRpYPtui8VvT6Q/edit

• Marinelli, Ryan Anthony & Eckhoff, Magnus Wiik (2025). Leaking LoRA. Full text in Research Archive Show summary

  Slides may be found at https://docs.google.com/presentation/d/1EEjD3wh9t3ZoneG1NBiXfv_0s3oyJbTYIQ0SygjnuJM/edit?usp=sharing This presentation was conducted to share research within the UiO community

• Marinelli, Ryan Anthony (2025). SAIL-K Framework For Secure AI Applications in Proceedings of NORA’s annual conference 2025. Nordic Machine Intelligence (NMI). doi: 10.5617/nmi.12551. Full text in Research Archive
• Marinelli, Ryan Anthony (2025). Practical IP Law For Research & Dependency Management. Full text in Research Archive Show summary

  Slides can be viewed at https://docs.google.com/presentation/d/1AU1VnGaxhFBnxupJQn4ABgn2h-TgCi7B7EFmhbcIZlM/edit?usp=sharing

• Marinelli, Ryan Anthony & Eckhoff, Magnus Wiik (2025). Leaking LoRa: An Evaluation of Password Leaks and Knowledge Storage in Large Language Models Storage in Large Language Model. Full text in Research Archive Show summary

  To effectively deploy Large Language Models (LLMs) in application-specific settings, fine-tuning techniques are applied to enhance performance on specialized tasks. This process often involves fine-tuning on user data data, which may contain sensitive information. Although not recommended, it is not uncommon for users to send passwords in messages, and fine-tuning models on this could result in passwords being leaked. In this study, a Large Language Model is fine-tuned with customer support data and passwords from the RockYou password wordlist using Low-Rank Adaptation (LoRA). Out of the first 200 passwords from the list, 37 were successfully recovered. Further, causal tracing is used to identify that password information is largely located in a few layers. Lastly, Rank One Model Editing (ROME) is used to remove the password information from the model, resulting in the number of passwords recovered going from 37 to 0. Link to preprint: https://arxiv.org/abs/2504.00031

• Marinelli, Ryan Anthony & Lammele, Markus (2025). Artificial Intelligence: A Blessing or a Curse Discussing Security Concerns of Diagnostic Models in Radiological Assessment. Full text in Research Archive Show summary

  The educational exhibit can be found at meeting proceedings https://www.asnr.org/wp-content/uploads/2025/05/ASNR25Proceedings_ExcerptaExhibits_05.01.25.pdf

• Marinelli, Ryan Anthony & Lammele, Markus (2024). Comparison of CNN Architecture and Policy Based Segmentation. Full text in Research Archive Show summary

  Poster located at https://drive.google.com/file/d/1Qsj1GU4Cgv623geK22cxMZVk61DjNyWa/view

• Marinelli, Ryan Anthony (2024). Hacking Machine Learning: Auditing Models Using Inversion Methods. Full text in Research Archive Show summary

  Event Information: https://www.meetup.com/ai-mad-lab/events/300788048/?recId=46e55b4c-6fba-4f09-8090-695aff110bb3&recSource=ml-popular-events-nearby&searchId=1e7d6020-8844-42bf-9653-ef0ed4258dc3 Slides: https://drive.google.com/file/d/1TPrk5jn9OaNDE_ZL4Q1UloOHoMVVp3VG/view?usp=sharing

• Marinelli, Ryan Anthony (2024). Artificial Intelligence: A Blessing or a Curse Discussing Security Concerns of Diagnostic Models in Radiological Assessment. Full text in Research Archive Show summary

  Abstract book for event is located at https://drive.google.com/file/d/1-cs85Ghl_UVEkqAak3EomUfpz3af3WXW/view?usp=sharing .

• Marinelli, Ryan Anthony; Erdodi, Laszlo Tibor; Sommervoll, Åvald Åslaugson & Zennaro, Fabio Massimo (2024). Function Approximation Techniques to Empower Complex Agents for SQLi. Full text in Research Archive Show summary

  Poster is located at https://drive.google.com/file/d/1kNmswe01EGW-Tg2tou6mToAb3NAnBPEf/view

• Marinelli, Ryan Anthony (2024). Natural Language Processing Seminar. Full text in Research Archive Show summary

  Guest lecture for data science students. Slides are located at https://docs.google.com/presentation/d/1-RWWem3GqzZyHqwAr_hHz7ptnBQnwEl2z6aqqxNB9fQ/edit?usp=sharing

• Cheng, Shen; Green, Seth & Marinelli, Ryan Anthony (2023). simpar: an R Package for Parameter Uncertainty Simulations in Pharmacometric Modeling. Full text in Research Archive Show summary

  The poster can be viewed at https://www.metrumrg.com/wp-content/uploads/2023/11/ACoP14_W-035_simpar-an-R-Package-for-Parameter-Uncertainty-Simulations-in-Pharmacometric-Modeling.pdf

• Chandler, David; Chandnani, Krisha; Goswami, Aditi; Marinelli, Ryan Anthony; Ness, Preben Monteiro & O’Donovan, Brian [Show all 20 contributors for this article] (2026). Data Study Group Final Report: CEFAS - Automated identification of sea pens using OpenCV and machine learning. The Alan Turing Institute. doi: https:/zenodo.org/records/20447259. Full text in Research Archive
• Adib, Shady; Bezalwar, Rashmi; Bolanos, Francisco; Brau, Wendy; Coghi, Francesco & Forshaw, Matthew [Show all 23 contributors for this article] (2026). Data Study Group Final Report: The Alan Turing Institute and Innovate UK BridgeAI - Understanding skill supply–demand gaps and recruitment trends for AI roles in the UK labour market. The Alan Turing Institute. doi: 10.5281/ZENODO.20607530. Full text in Research Archive Show summary

  Data Study Groups are week-long events at The Alan Turing Institute bringing together some of the country’s top talent from data science, artificial intelligence, and wider fields, to analyse real-world data science challenges. This report was produced as part of The Alan Turing Institute Data Study Group in collaboration with the Innovate UK BridgeAI programme [6]. The Innovate UK BridgeAI programme is an Innovate UK initiative that aims to accelerate the adoption of artificial intelligence (AI) across key UK sectors by supporting skills development, workforce readiness, and responsible AI deployment. It addresses a central challenge in AI skills policy: how to connect structured competency frameworks and occupational standards with fast-changing labour market evidence from job advertisements and macro-level labour market indicators. This analysis focuses on three related questions: how AI-related roles and skills can be identified and aligned across Innovate UK BridgeAI, Skills England, and labour market data; how AI-related hiring demand relates to broader labour market conditions and whether vacancy data improve macroeconomic monitoring and forecasting; and how job-posting evidence can support occupational progression analysis and prototype AI job classification.

• Marinelli, Ryan Anthony (2026). Fortifying AI in Deployment Contexts: A Defense in Depth Approach. Universitetet i Oslo. Full text in Research Archive Show summary

  The underlying goal of this project is to aid in fortifying productionized AI- powered applications. Previously, CRUD (Create, Read, Update, Delete) web applications were the norm with users interacting only with application logic and the database. However, with the greater proliferation of machine learning being intertwined into core functionality of applications, there is a greater demand to secure not only the conventional architecture, but AI components that are becoming an expected part of application and system architecture. In this pursuit, a defense-in-depth approach is taken by first segmenting AI applications into meaningful layers of analysis and attempting to secure each layer, promoting a robust strategy to support the deployment of modern applications with increasing operational complexity.

View all works in NVA