Richard Hladík

[CV] [Google Scholar] [Github] [Youtube] [Research]

I am a second-year PhD student at ETH Zurich, fortunate to be advised by Bernhard Haeupler.

In my free time, I like playing music, wandering around in the mountains, and organizing programming summer camps for highschoolers.

With my friends, we also run Polylog, a Youtube channel where we explain various computer science topics.

Other

• Impromptu – a party game revolving around creating AI-generated images, making up funny prompts for them, and trying to trick others. Created at the time when AI image generation was still kind of wonky. It didn’t catch on as much as we hoped, but it was a fun project. Unmaintained since 2023, but still working and apparently there are still people playing it to this day.
• Outotune – a VST plugin that allows you to sing chords in real time using your voice and a keyboard. Inspired by Jacob Collier’s harmonizer. I kind of abandoned the project after I got a first working prototype, but one day I’m going to come back and make this into a full-fledged plugin.

Research

I’m primarily interested in graph theory, algorithms and data structures.

Publications

• Fast and Simple Sorting Using Partial Information – Algorithmica, May 2026 (Conference version: SODA 2025)
  with: Bernhard Haeupler, John Iacono, Václav Rozhoň, Robert E. Tarjan, Jakub Tětek
  About optimally solving the 1976 problem of sorting under partial information the combination of a simple algorithm and beyond-worst-case heaps.
  [arXiv][PDF][slides][DOI][conference version]
• Stronger Directed Low-Diameter Decompositions with Sub-Logarithmic Diameter and Separation – 2025 (preprint)
  with: Bernhard Haeupler, Shengzhe Wang, Zhijun Zhang
  Strenghtens the guarantees of two state-of-the-art algorithms for directed low-diameter decompositions.
  [arXiv][PDF]
• Near-Universally-Optimal Differentially Private Minimum Spanning Trees – FORC 2025, Best student paper award
  with: Jakub Tětek
  About the problem of privately releasing the minimum spanning trees of the graph, and how one can use combinatorial properties of spanning trees to design lower bounds for this problem.
  [arXiv][PDF][talk][slides][DOI]
• Smooth Sensitivity Revisited: Towards Optimality – FORC 2025
  with: Jakub Tětek
  Introduces a new distribution that improves accuracy in privacy-preserving computations via smooth sensitivity.
  [arXiv][PDF]
• Bidirectional Dijkstra's Algorithm is Instance-Optimal – SOSA 2025
  with: Bernhard Haeupler, Václav Rozhoň, Robert E. Tarjan, Jakub Tětek
  About how bidirectional Dijkstra's algorithm (also known as an example of the more general meet in the middle technique) is optimal on every graph, in the sense of instance-optimality.
  [arXiv][PDF][slides][DOI]
• Fast and Simple Sorting Using Partial Information – Diploma thesis, ETH Zurich, 2024, ETH medal for an outstanding Master's thesis
  A streamlined version of the SODA 2025 paper of the same name, with the algorithm and proof fitting into 5.5 pages.
  [link]
• Universal Optimality of Dijkstra via Beyond-Worst-Case Heaps – FOCS 2024, Best paper award
  with: Bernhard Haeupler, Václav Rozhoň, Robert E. Tarjan, Jakub Tětek
  About how certain implementations of Dijkstra are in some sense optimal on every graph, if one uses a nice enough heap. Featured in Quanta Magazine and Wired.
  [arXiv][PDF][talk][slides][DOI]
• Combinatorial Algorithms for Flow Problems – Bachelor's thesis, Charles University, 2021
  My bachelor's thesis on the multicommodity flow problem and why it is so hard to find a combinatorial algorithm for it. Supervised by Martin Koutecký and one of the reasons why I decided to go into research.
  [link]
• On the Complexity of a Periodic Scheduling Problem with Precedence Relations – COCOA 2020
  with: Anna Minaeva, Zdeněk Hanzálek
  About a certain periodic scheduling problem, its NP-hardness, and heuristics that seem to work well in practice.
  [DOI]