Jakub Tětek

Looking for summer interns, Master’s students, and advanced undergrads to join my #DifferentialPrivacy research at @INSAITinstitute! Interested? Email me your CV and a short note on what specifically excites you in the area of differential privacy.

Honored to share that our paper on the universal optimality of Dijkstra's algorithm received the Best Paper Award at FOCS 2024! Grateful to co-authors Bernhard Haeupler, Richard Hladík, Václav Rozhoň, and Bob Tarjan. Read more here: arxiv.org/abs/2311.11793

A new proof shows that an upgraded version of a 70-year-old algorithm reigns supreme: It finds the most efficient pathways through any graph. @benbenbrubaker reports: quantamagazine.org/computer-scien…

I am very excited to announce that I will be joining the brand new theory group of Amir Abboud at @INSAITinstitute as a postdoc. At first, I wanted to go for a tenure-track position in my home country, but this opportunity was just too good to turn down :-)

Universal optimality is a strong beyond-worst-case optimality guarantee for algorithms on weighted graphs, commonly used in distributed algorithms. Our paper gives the first use of universal optimality in differential privacy, namely for releasing MST: arxiv.org/abs/2404.15035

BARC 2 starts in 2024. This celebration commemorates the success of BARC 1. Bob Tarjan will be speaking at the event. Please register and join, if you are in the neighbourhood. di.ku.dk/english/event-…

The talk about our paper on the instance-optimality of Dijkstra's algorithm is now online! Check it out here: youtu.be/z_hkBZ0lDFU

A new and fundamental instance-optimality result 🤯 (Who knew there was still more to prove about Dijkstra's algorithm?)

If you are interested in doing a #PhD in #algorithms #theory, or more generally in theoretical computer science, you should consider Copenhagen and @barcdk! We strive to do deep and impactful work in a fun and collaborative atmosphere.

Proud of @barcdk PhD students Christian Lebeda and Jakub Tětek who managed to crack one of my favorite problems in differentially private streaming: How to get good utility out of deterministic heavy hitters sketches. Their paper is at arxiv.org/abs/2301.02457 but a short summary: