iiinova — where your digital self keeps evolving

Arno Solin @arnosolin.bsky.social · 08/10/2026

🔥 #LeAVJEPA: A simple recipe for self-supervised learning from sound and vision 🔥 One shared encoder, trained without class labels. Dropping a modality makes audio & video learn a common representation. 🧵

240

Arno Solin @arnosolin.bsky.social · 25/06/2026

6/ Smol-GS is not just smaller. It keeps rendering fast, preserves high visual quality, and avoids the structural overhead of anchors. 🥇On the 3DGS.zip benchmark, it currently gives the best size–quality trade-off: w-m.github.io/3dgs-compres...

100

Arno Solin @arnosolin.bsky.social · 25/06/2026

5/ Tiny MLPs decode each splat into what the renderer needs: color, opacity, shape, rotation, and view-dependent effects. The features and scale controllers are quantized and entropy-coded. The result is a compact neural splat representation without anchor-offset bookkeeping.

110

Arno Solin @arnosolin.bsky.social · 25/06/2026

4/ In Smol-GS, each splat has a compact learned feature vector. Geometry stays explicit, but coordinates are stored with an occupancy-octree. The same octree also gives a locality-aware positional encoding. In short: explicit geometry + learned local appearance.

100

Arno Solin @arnosolin.bsky.social · 25/06/2026

3/ Most strong 3DGS compression methods are anchor-based: Store shared anchors, use local offsets, and decode splats around them. That works well and has dominated recent benchmarks. But Smol-GS takes a different approach.

100

Arno Solin @arnosolin.bsky.social · 25/06/2026

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.

170

Arno Solin @arnosolin.bsky.social · 14/04/2026

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.

291

Arno Solin @arnosolin.bsky.social · 28/11/2025

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

1136

Arno Solin @arnosolin.bsky.social · 12/08/2025

📣 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]

23721

Arno Solin @arnosolin.bsky.social · 09/06/2025

Qualitative visualization of static distractor elements achieved by our model, DeSplat. [3/n]

100

Arno Solin @arnosolin.bsky.social · 09/06/2025

Compared to Splatfacto we model and can ignore distractors to improve 3DGS reconstruction quality. [2/n]

100

Arno Solin @arnosolin.bsky.social · 09/06/2025

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]

150

Arno Solin @arnosolin.bsky.social · 04/06/2025

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.

0162

Arno Solin @arnosolin.bsky.social · 29/04/2025

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]

34411

Arno Solin @arnosolin.bsky.social · 27/04/2025

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

170

Arno Solin @arnosolin.bsky.social · 23/04/2025

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

070

Arno Solin @arnosolin.bsky.social · 23/04/2025

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

151

Arno Solin @arnosolin.bsky.social · 23/04/2025

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

1110

Arno Solin @arnosolin.bsky.social · 21/04/2025

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

2113

Arno Solin @arnosolin.bsky.social · 21/04/2025

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

0101

Arno Solin @arnosolin.bsky.social · 21/04/2025

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.

130

Arno Solin @arnosolin.bsky.social · 08/03/2025

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.

2136

Arno Solin @arnosolin.bsky.social · 15/01/2025

I also stumbled upon the “Classical Mechanics Lab” (in the picture), and physics labs sure are different than I thought.

050

Arno Solin @arnosolin.bsky.social · 15/01/2025

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. ✨

1170

Arno Solin @arnosolin.bsky.social · 15/12/2024

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

0402

Arno Solin @arnosolin.bsky.social · 11/12/2024

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

0202