Mykola Lukashchuk
@mykolalukashchuk.bsky.social
Phd Candidate, BIASlab, TU/e
Freenergian
Freenergian
Reposted by Mykola Lukashchuk
The invited NeurIPS talk by Zeynep Tufekci. Very interesting'
December 3, 2025 at 11:11 PM
The invited NeurIPS talk by Zeynep Tufekci. Very interesting'
Reposted by Mykola Lukashchuk
In this paper with my amazing colleague we tried to attack probability foundations from the metric space theory perspective.
November 25, 2025 at 1:33 PM
In this paper with my amazing colleague we tried to attack probability foundations from the metric space theory perspective.
Reposted by Mykola Lukashchuk
Wouter W. L. Nuijten, Mykola Lukashchuk: Active Inference is a Subtype of Variational Inference https://arxiv.org/abs/2511.18955 https://arxiv.org/pdf/2511.18955 https://arxiv.org/html/2511.18955
November 25, 2025 at 6:29 AM
Wouter W. L. Nuijten, Mykola Lukashchuk: Active Inference is a Subtype of Variational Inference https://arxiv.org/abs/2511.18955 https://arxiv.org/pdf/2511.18955 https://arxiv.org/html/2511.18955
Reposted by Mykola Lukashchuk
Rapha\"el Tr\'esor, Mykola Lukashchuk: Resolution of the Borel-Kolmogorov Paradox via the Maximum Entropy Principle https://arxiv.org/abs/2509.24735 https://arxiv.org/pdf/2509.24735 https://arxiv.org/html/2509.24735
September 30, 2025 at 6:41 AM
Rapha\"el Tr\'esor, Mykola Lukashchuk: Resolution of the Borel-Kolmogorov Paradox via the Maximum Entropy Principle https://arxiv.org/abs/2509.24735 https://arxiv.org/pdf/2509.24735 https://arxiv.org/html/2509.24735
Reposted by Mykola Lukashchuk
Grinding the Silksong is a great self-reflection instrument
November 23, 2025 at 11:47 AM
Grinding the Silksong is a great self-reflection instrument
Reposted by Mykola Lukashchuk
Reposted by Mykola Lukashchuk
Kyrylo Yemets, Mykola Lukashchuk, Ivan Izonin: GateTS: Versatile and Efficient Forecasting via Attention-Inspired routed Mixture-of-Experts https://arxiv.org/abs/2508.17515 https://arxiv.org/pdf/2508.17515 https://arxiv.org/html/2508.17515
August 26, 2025 at 6:34 AM
Kyrylo Yemets, Mykola Lukashchuk, Ivan Izonin: GateTS: Versatile and Efficient Forecasting via Attention-Inspired routed Mixture-of-Experts https://arxiv.org/abs/2508.17515 https://arxiv.org/pdf/2508.17515 https://arxiv.org/html/2508.17515
Reposted by Mykola Lukashchuk
Introducing TidierErrors.jl, an experimental #JuliaLang package that makes error messages tidier with the help of LLMs (with support for local LLMs!)
github.com/TidierOrg/Ti...
Developed by @randy.pub
Left: original error message
Right: tidier error message
github.com/TidierOrg/Ti...
Developed by @randy.pub
Left: original error message
Right: tidier error message
August 11, 2025 at 3:10 AM
Introducing TidierErrors.jl, an experimental #JuliaLang package that makes error messages tidier with the help of LLMs (with support for local LLMs!)
github.com/TidierOrg/Ti...
Developed by @randy.pub
Left: original error message
Right: tidier error message
github.com/TidierOrg/Ti...
Developed by @randy.pub
Left: original error message
Right: tidier error message
Reposted by Mykola Lukashchuk
Fixed it for you...
June 18, 2025 at 4:59 PM
Fixed it for you...
Reposted by Mykola Lukashchuk
Wouter W. L. Nuijten, Mykola Lukashchuk, Thijs van de Laar, Bert de Vries: A Message Passing Realization of Expected Free Energy Minimization https://arxiv.org/abs/2508.02197 https://arxiv.org/pdf/2508.02197 https://arxiv.org/html/2508.02197
August 5, 2025 at 6:30 AM
Wouter W. L. Nuijten, Mykola Lukashchuk, Thijs van de Laar, Bert de Vries: A Message Passing Realization of Expected Free Energy Minimization https://arxiv.org/abs/2508.02197 https://arxiv.org/pdf/2508.02197 https://arxiv.org/html/2508.02197
If you're running optimization procedures and struggling with learning rate tuning, I found a simple, effective method that works with any optimizer: DoG arxiv.org/pdf/2302.12022. The formula is elegant and easily generalizes to Riemannian manifolds.
arxiv.org
August 4, 2025 at 9:29 AM
If you're running optimization procedures and struggling with learning rate tuning, I found a simple, effective method that works with any optimizer: DoG arxiv.org/pdf/2302.12022. The formula is elegant and easily generalizes to Riemannian manifolds.
Ah definitely this one is the last one.
May 16, 2025 at 10:29 AM
Ah definitely this one is the last one.
Reposted by Mykola Lukashchuk
It is alive! It is alive!
May 14, 2025 at 10:05 AM
It is alive! It is alive!
Reposted by Mykola Lukashchuk
Bad news: Overleaf is down. The only way to edit LaTeX known to humankind!
May 14, 2025 at 7:27 AM
Bad news: Overleaf is down. The only way to edit LaTeX known to humankind!
Reposted by Mykola Lukashchuk
Reposted by Mykola Lukashchuk
A new nice paper from my lab.
Bert de Vries, Wouter Nuijten, Thijs van de Laar, Wouter Kouw, Sepideh Adamiat, Tim Nisslbeck, Mykola Lukashchuk, Hoang Minh Huu Nguyen, Marco Hidalgo Araya, Raphael Tresor, Thijs Jenneskens, ...
Expected Free Energy-based Planning as Variational Inference
https://arxiv.org/abs/2504.14898
Expected Free Energy-based Planning as Variational Inference
https://arxiv.org/abs/2504.14898
April 25, 2025 at 3:29 PM
A new nice paper from my lab.
My friends at Lazy Dynamics just launched Python support for RxInfer models! 🚀
Check it out: lazydynamics.github.io/RxInferClien...
Julia + Fast Bayesian Inference + Python = ¿Por qué no los tres?
P.S. Maybe this will finally get them on Bluesky... 😂
Check it out: lazydynamics.github.io/RxInferClien...
Julia + Fast Bayesian Inference + Python = ¿Por qué no los tres?
P.S. Maybe this will finally get them on Bluesky... 😂
Inference in State Space Models - RxInfer Client
Python client for the RxInfer API
lazydynamics.github.io
April 14, 2025 at 8:37 AM
My friends at Lazy Dynamics just launched Python support for RxInfer models! 🚀
Check it out: lazydynamics.github.io/RxInferClien...
Julia + Fast Bayesian Inference + Python = ¿Por qué no los tres?
P.S. Maybe this will finally get them on Bluesky... 😂
Check it out: lazydynamics.github.io/RxInferClien...
Julia + Fast Bayesian Inference + Python = ¿Por qué no los tres?
P.S. Maybe this will finally get them on Bluesky... 😂