# Pasquale Minervini

> Official personal academic website of Dr Pasquale Minervini, a researcher and faculty member at the University of Edinburgh, co-founder and CTO at Miniml.AI, and ELLIS Scholar. The site covers research in natural language processing, machine learning, large language models, knowledge representation and reasoning, robustness, efficiency, and neuro-symbolic AI.

Use the Publications page for the site-maintained list of papers and BibTeX links. Google Scholar provides a more exhaustive and frequently updated publication history. Blog posts describe research results, tools, and opportunities and should be interpreted using their publication dates.

## Core resources

- [Homepage](https://neuralnoise.com/): Biography, current roles, and the latest posts.
- [Publications](https://neuralnoise.com/publications/): Research grants, patents, tutorials, and publications with BibTeX links.
- [Curriculum vitae](https://neuralnoise.com/documents/CV.pdf): Detailed academic and professional history in PDF form.
- [Google Scholar](https://scholar.google.com/citations?hl=en&user=9sk6CSgAAAAJ&view_op=list_works&sortby=pubdate): More exhaustive and frequently updated publication record.

## Selected research and technical writing

- [{ICML, ACL} 2026](https://neuralnoise.com/2026/icml-acl-2026/): Current conference work on GUI agents, efficient generation, reasoning, robustness, and mechanistic interpretability.
- [Real-World Impact of Our Research](https://neuralnoise.com/2025/research-impact/): Examples of research adoption and impact in industry.
- [Some Notes on Gradient Estimation](https://neuralnoise.com/2026/some-notes-on-gradient-estimation/): Technical notes on gradient estimators for stochastic and discrete objectives.
- [Deep Research MCP](https://neuralnoise.com/2026/deep-research-mcp/): A tool exposing multiple deep-research backends through the Model Context Protocol.

## Optional

- [GitHub profile](https://github.com/pminervini): Open-source projects and code.
- [RSS feed](https://neuralnoise.com/feed.xml): The ten most recent site posts.
- [Sitemap](https://neuralnoise.com/sitemap.xml): Complete index of published site pages and posts.