Rasmus Pagh
banner
rasmuspagh.net
Rasmus Pagh
@rasmuspagh.net
Professor of computer science at University of Copenhagen. Interested in random things & their application (especially to algorithms and privacy). rasmuspagh.net
The papers on the DL are open access starting this year, so I think there is good movement towards open access. Making some metadata paywalled is a setback but hopefully it can be reversed.
January 9, 2026 at 9:08 AM
Reposted by Rasmus Pagh
Starting in the fall (Gemini 2.5 pro launched experimentally at the end of March) LLMs that were useful for math launched. I didn't have much luck using them within their web interface, but I found them to be extremely useful within coding agent harnesses like Windsurf.
December 21, 2025 at 7:01 PM
"use of generative AI tools and technologies to create content is permitted but must be fully disclosed in the Work [..] Basic word processing systems that recommend and insert replacement text, perform spelling or grammar checks [..] are to be considered exceptions to this disclosure requirement"
October 27, 2025 at 10:59 AM
Agreed, if what we are seeing is a lot of AI-generated, mediocre applications being added to the pool it won't mean much for the chances (but additional work for the panels...). I guess time will tell.
September 4, 2025 at 2:24 PM
VLDB 2025 - Conference Awards
Best Paper, Best Demo, and Distinguished Reviewers.
vldb.org
September 4, 2025 at 6:48 AM
Good question. I would be curious to know if applications from researchers in the United States has significantly increased following the decrease in funding there?
September 2, 2025 at 3:10 PM
The Open Diffix synthetic data generator system can be found at www.open-diffix.org
Open Diffix
Strong Anonymization for Structured Data. Open. Free.
www.open-diffix.org
September 2, 2025 at 12:42 PM
I wonder if there are independent attempts to attack the protections of this software and/or quantify its utility? (A prior version of Diffix was famously broken by Aloni Cohen and Kobbi Nissim, @kobbini.bsky.social, in arxiv.org/abs/1810.05692)
Linear Program Reconstruction in Practice
We briefly report on a successful linear program reconstruction attack performed on a production statistical queries system and using a real dataset. The attack was deployed in test environment in the...
arxiv.org
September 2, 2025 at 12:41 PM