3D Gaussian splats are amazing--until you realise how much storage space they need. Smol-GS makes #3DGS actually small: explicit splats, tiny learned features w/ positional awareness, octree-coded geometry, and fast rendering. 🥇Now leading the 3DGS.zip benchmark.

👋🇧🇷 If you are at #ICLR2026 today, you should talk to @antonbaumann.bsky.social who is presenting our paper about turning pre-trained VLMs into probabilistic models without retraining or fine-tuning. Poster Session 3 ⌚: 10:30am - 1:00pm (local time) 📍: Pavilion 3 P3 - #313 @iclr-conf.bsky.social

1/ 🔥 New paper: Differentiable Vector Quantization (DiVeQ) 🔥 Vector quantization (VQ) is a core tool in modern AI. It connects continuous data like images and audio to discrete tokens used by transformers. It underpins compression, generation, and multimodal modelling.

Statement from #AISTATS2026 organizers regarding the @openreview.bsky.social API Security Incident

I recently gave my installation talk after being tenured. The video of the talk is now available on the university's YouTube channel: youtu.be/R1UQoflPTDg 1/n

Making sense of learning machines – Arno Solin

YouTube video by Aalto University

youtu.be

📣 Please share: We invite submissions to the 29th International Conference on Artificial Intelligence and Statistics (#AISTATS 2026) and welcome paper submissions at the intersection of AI, machine learning, statistics, and related areas. [1/3]

Remember that computers use bitstrings to represent numbers? We exploit this in our recent @auai.org paper and introduce #BitVI. #BitVI directly learns an approximation in the space of bitstring representations, thus, capturing complex distributions under varying numerical precision regimes.

Real-world #3DGS scenes are messy—occluders, moving objects, and clutter often ruin reconstruction. This #CVPR2025 paper presents DeSplat, which separates static scene content from distractors, all without requiring external semantic models. [1/n]

I’m visiting the Isaac Newton Institute for Mathematical Sciences in Cambridge this week. I’m giving an invited talk in the ”Calibrating prediction uncertainty : statistics and machine learning perspectives” workshop on Thursday.

Have you thought that in computer memory model weights are given in terms of discrete values in any case. Thus, why not do probabilistic inference on the discrete (quantized) parameters. @trappmartin.bsky.social is presenting our work at #AABI2025 today. [1/3]

Excited to share "Plan*RAG: Efficient Test-Time Planning for Retrieval Augmented Generation", presented at the #ICLR2025 "Workshop on Reasoning and Planning for LLMs" on Monday! 🚀 1/3

Our TMLR-to-ICLR poster "Exploiting Hankel-Toeplitz Structures for Fast Computation of Kernel Precision Matrices" (Frida Viset, Anton Kullberg, Frederiek Wesel, Arno Solin) 🗓️ Hall 3 + Hall 2B #416, Fri 25 Apr 10 a.m. +08 — 12:30 p.m. +08 📄 Preprint: arxiv.org/abs/2408.02346

Our #ICLR2025 poster "Equivariant Denoisers Cannot Copy Graphs: Align Your Graph Diffusion Models" (Najwa Laabid, Severi Rissanen, Markus Heinonen, Arno Solin, Vikas Garg) 🗓️ Hall 3 + Hall 2B #194, Fri 25 Apr 3 p.m. +08 — 5:30 p.m. +08 📄 Preprint: arxiv.org/abs/2405.17656

Our #ICLR2025 poster "Streamlining Prediction in Bayesian Deep Learning" (Rui Li · Marcus Klasson, Arno Solin, Martin Trapp) 🗓️ Hall 3 + Hall 2B #413, Fri 25 Apr 10 a.m. +08 — 12:30 p.m. +08 📄 Preprint: arxiv.org/abs/2411.18425

Our #ICLR2025 poster "Discrete Codebook World Models for Continuous Control" (Aidan Scannell, Mohammadreza Nakhaeinezhadfard, Kalle Kujanpää, Yi Zhao, Kevin Luck, Arno Solin, Joni Pajarinen) 🗓️ Hall 3 + Hall 2B #415, Thu 24 Apr 10 a.m. +08 — 12:30 p.m. +08 📄 Preprint: arxiv.org/abs/2503.00653

Our #ICLR2025 poster "Free Hunch: Denoiser Covariance Estimation for Diffusion Models Without Extra Costs" (Severi Rissanen, Markus Heinonen, Arno Solin) 🗓️ Hall 3 + Hall 2B #140, Thu 24 Apr 3 p.m. +08 — 5:30 p.m. +08 📄 Preprint: arxiv.org/abs/2410.11149

This week, we are presenting five papers at the main conference of the Thirteenth International Conference on Learning Representations (#ICLR2025) in Singapore. You can find my research group members and collaborators at the following posters.

There is still time to submit your papers to our #CVPR2025 workshop on Uncertainty Quantification for Computer Vision, which is part of the workshop lineup at the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) in Nashville, Tennessee.

Are you going to be at #WACV and want to know if “Flatness Improves Backbone Generalisation in Few-shot Classification”? Then join the oral presentation by @ruili-pml.bsky.social of our paper! 🔗 lnkd.in/dBMmN7Vs Done together with @marcusklasson.bsky.social and @arnosolin.bsky.social.

This week I have been teaching ML outside my own academic bubble. 🫧 I have been giving a crash course as part of the #Nordita Winter School on "Physics of Machine Learning & Machine Learning for Physics“ in Stockholm. Great interaction with young physicists and new avenues for applying ML. ✨

🔥Great workshop! — #NeurIPS workshop on Bayesian Decision-making and Uncertainty. \w @trappmartin.bsky.social

On Wednesday, Oliver Hamelijnck (with Theo Damoulas and me) is presenting our paper "Physics-Informed Variational State-Space Gaussian Processes" at NeurIPS in Vancouver. Come by poster 4204 in East Exhibit Hall A-C starting 11 am. 📝 Paper pre-print: arxiv.org/abs/2409.13876

I will present ✌️ BDU workshop papers @ NeurIPS: one by Rui Li (looking for internships) and one by Anton Baumann. 🔗 to extended versions: 1. 🙋 "How can we make predictions in BDL efficiently?" 👉 arxiv.org/abs/2411.18425 2. 🙋 "How can we do prob. active learning in VLMs" 👉 arxiv.org/abs/2412.06014

Post-hoc Probabilistic Vision-Language Models

Vision-language models (VLMs), such as CLIP and SigLIP, have found remarkable success in classification, retrieval, and generative tasks. For this, VLMs deterministically map images and text descripti...

arxiv.org

Martin Trapp@trappmartin.eurosky.social · 2y ago

On my way to #NeurIPS. Looking forward to seeing many friends again. Ping me if you want to chat, always happy to meet new people. :)

I will be at #NeurIPS2024 in Vancouver. I’m looking for post-docs, and if you want to talk about post-doc opportunities, get in touch. 🤗 Here’s my current team at Aalto University: users.aalto.fi/~asolin/group/

DeSplat: Decomposed Gaussian Splatting for Distractor-Free Rendering Yihao Wang, Marcus Klasson, Matias Turkulainen, Shuzhe Wang, Juho Kannala, @arnosolin.bsky.social tl;dr: decompose alpha compositing and explicitly separate occluders and the underlying static 3D scene arxiv.org/abs/2411.19756