PhD candidate at MIMUW, Head of AI at TradeLink, member of PL4AI, motorcycle and bike enthusiast.
Warsaw, Poland
Joined May 2017
- Tweets96
- Following162
- Followers86
- Likes295
Pinned Tweet
Our paper "A Wiener process perspective on local intrinsic dimension estimation methods" has been accepted to AAAI. Great thanks to Adam Kurpisz , Łukasz Garncarek and @d__filipiak arxiv.org/abs/2406.17125
Piotr Tempczyk retweeted
Shoutout to @PiotrTempczyk et al. (authors of LIDL) for releasing their new work (arxiv.org/abs/2406.17125) and enriching FLIPD shortly after it came out!
Piotr Tempczyk retweeted
🚀If you're interested in more manifold learning, we released another paper on the first differentiable intrinsic dimension estimator using diffusion models!
👨🏫I will present it at 2 orals @icmlconf workshops
Differentiable Almost Everything, and SPIGM.
Follow for more! 1/n 🧵
🚀 Excited to present our work "A Geometric Explanation of the Likelihood OOD Detection Paradox", a paradox first introduced by @eric_nalisnick et al., at the #ICML2024 conference.
📔icml.cc/virtual/2024/poster/…
Follow the thread to find more! 1/8 🧵
@VectorInst @Layer6AI @UofT
Piotr Tempczyk retweeted
If you're attending @icmlconf, please drop by our poster at 1:30 pm:
"A geometric explanation of the likelihood OOD detection paradox"
Hall C 4-9 #417
@JStanczuk, @eric_nalisnick
Piotr Tempczyk retweeted
After sharing my thoughts on LLMs in the morning, now we have @jesfrellsen introducing energy-based models at the best #GenAI summer school (gemss.ai/2024/) at @TUeindhoven, at the heart of @Brainport_int!
Happy to announce that a draft of our paper with Adam Kurpisz, Łukasz Garncarek and @d__filipiak exploring theoretical properties of intrinsic dimension estimation algorithms (LIDL, FLIPD) is now online: arxiv.org/abs/2406.17125 @przem_spurek @adam_golinski @hamid_R_kamkar
It is a great pleasure to be here, wonderful event and great people!
We have kicked off! The best summer school on #GenerativeAI: ✨Generative Modeling Summer School 2024✨! We turn @TUeindhoven into the world's capitol of #GenAI. Amazing lineup: @jesfrellsen, @pamattei, @dpkingma, @liyzhen2, @helibenhamu, @StephanMandt, @ewa_szczurek, @vlamen 🤩
Piotr Tempczyk retweeted
"Better than what's known" - the dream of AI in model-based optimization. Excited to present my paper with @pabbeel & @svlevine on Functional Graphical Models (arxiv.org/pdf/2401.05442) giving theoretical foundations for chasing that dream, at @aistats_conf ! #aistats_conf #ai – at Valencia, Spain
Piotr Tempczyk retweeted
Deep neural nets with the right symmetries baked-in (e.g. translational equivariance in CNNs) perform better. But can we learn them from data? We can with differentiable Laplace approximations! 🌟New method: LILA🌟 with @a1mmer @vincefort, @gxr, @markvanderwilk. A thread👇 🧵1/10
Piotr Tempczyk retweeted
Vincent is an amazing collaborator. If you're interested in probabilistic/Bayesian ML, you should definitely take a look!
I'm beyond excited to announce that I will be starting my own research group @helmholtz_ai in Munich! I'm hiring PhD students and postdocs, so please share these opportunities widely within your circles. Details in thread.
Piotr Tempczyk retweeted
The ProbAI 2022 materials 📚 and recorded lectures 🎥 are all available to the public. github.com/probabilisticai/p…
Piotr Tempczyk retweeted
Deadline is around two weeks! Please submit your 4-page extended abstracts on CMT!
Announcing the 2022 NeurIPS Workshop on Gaussian Processes, Spatiotemporal Modeling, and Decision-making Systems!
Please consider submitting a 4-page extended abstract by September 22nd!
gp-seminar-series.github.io/…
Thanks to amazing organizers @gpleiss, @liza_p_semenova, @ziwphd
ALT NeurIPS Workshop on Gaussian Processes, Spatiotemporal Modeling, and Decision-making Systems
My presentation of LIDL paper from #ICML2022 is now online: slideslive.com/38983065/lidl… @adam_golinski @przem_spurek @r_michaluk – at Warsaw, Poland
Piotr Tempczyk retweeted
"Cat Knight".
#StableDiffusion #AI
Prompt: kneeling cat knight, portrait, finely detailed armor, intricate design, silver, silk, cinematic lighting, 4k
Parameters:scale7.50-k euler
By Valere: hostux.social/@valere/108939…
Piotr Tempczyk retweeted
"Voyage through Time"
is my first artpiece using #stablediffusion and I am blown away with the possibilities...
We're crossing a threshold where generative AI is no longer just about novel aesthetics, but evolving into an amazing tool to build powerful, human-centered narratives
Piotr Tempczyk retweeted
The AISTATS 2023 Call for Papers is out!
aistats.org/aistats2023/call…
Submission Deadline: Thu Oct 13, AOE
Program Chairs: Jennifer Dy, Jan-Willem van de Meent (@jwvdm)
General Chair: Francisco Ruiz
Workflow Chairs: Christian Naesseth (@chris_naesseth), Davin Hill (@davinhill_)
Piotr Tempczyk retweeted
It was a pleasure and an honour to speak about some recent advances in Bayesian spatial modelling. Many thanks to the organisers for the invitation and their hard work @tetraduzione @martin_trapp @fabri_ven
#TPM2022 #UAI2022
@UncertaintyInAI
@liza_p_semenova on how #geospatial #probabilistic reasoning can make us into modern Jon Snow, the father of #epidemiology ...and #inference!
#TPM2022 #UAI2022 @UncertaintyInAI
Piotr Tempczyk retweeted
We propose a numerically stable conjugate gradients in half precision, enabling Gaussian processes on 1.8 million points on a single GPU in 10 hours, without any sparse approximations!
arxiv.org/abs/2207.06856
Our new #UAI2022 paper with Andres Potapczynski and Wesley Maddox. 1/2
Piotr Tempczyk retweeted
#TPM2022 just ended
thanks everyone and @jmtomczak for the nice final discussion and...see you soon!
@UncertaintyInAI
Want to know more about our work on “Learning Invariant Weights in NNs”? We show how symmetry properties can be learned with gradients. Chat to us at #uai2022 poster #419 tomorrow, 3 Aug.
Oral presentation slides: tychovdo.github.io/slides/ua…. Paper: arxiv.org/abs/2202.12439.
