Lightnews — Scholar-powered news

Arno Solin @arnosolin.bsky.social · Aug 12

Yes. The easiest way to find it will be on the website virtual.aistats.org We are in the process of adding material there and will add a link.

2026 Conference

virtual.aistats.org

1

Arno Solin @arnosolin.bsky.social · Aug 12

We will go public with it as soon as everything is set up with the venue.

1 3

Arno Solin @arnosolin.bsky.social · Aug 12

I'm thrilled to be Program Chairing AISTATS 2026 together with Aaditya Ramdas. AISTATS has a special feel to it, and it has been described by many colleagues as their "favourite conference". We aim to preserve that spirit while introducing some fresh elements for 2026. [3/3]

1 4

Arno Solin @arnosolin.bsky.social · Aug 12

Accepted papers will be presented in person in Morocco, May 2–5, 2026. The full Call for Papers is available here: virtual.aistats.org/Conferences/... [2/3]

Call for Papers

virtual.aistats.org

1 1

Arno Solin @arnosolin.bsky.social · Aug 12

📣 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]

2 21 36

Reposted by Arno Solin

Martin Trapp @trappmartin.bsky.social · Jul 21

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.

2 3 22

Arno Solin @arnosolin.bsky.social · Jun 9

Check our #CVPR paper and project page for more results, videos, and code!
📄 arxiv.org/abs/2411.19756
🎈 aaltoml.github.io/desplat/

DeSplat: Decomposed Gaussian Splatting for Distractor-Free Rendering

aaltoml.github.io

Arno Solin @arnosolin.bsky.social · Jun 9

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

1

Arno Solin @arnosolin.bsky.social · Jun 9

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

1

Arno Solin @arnosolin.bsky.social · Jun 9

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]

1 5

Arno Solin @arnosolin.bsky.social · Jun 4

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.

2 16

Arno Solin @arnosolin.bsky.social · Apr 29

Our method addresses the eminent question of probabilistic modelling in quantized large-scale ML models. See the workshop paper below. [3/3]

📄 Paper: openreview.net/forum?id=Sai...

Are Your Continuous Approximations Really Continuous? Reimagining...

Efficiently performing probabilistic inference in large models is a significant challenge due to the high computational demands and continuous nature of the model parameters. At the same time, the...

openreview.net

1

Arno Solin @arnosolin.bsky.social · Apr 29

We introduce BitVI, a novel approach for variational inference with discrete bitstring representations of continuous parameters. We use a deterministic probabilistic circuit structure to model the distribution over bitstrings, allowing for exact and efficient probabilistic inference. [2/3]

1 1

Arno Solin @arnosolin.bsky.social · Apr 29

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]

3 11 46

Arno Solin @arnosolin.bsky.social · Apr 27

We show that externalising reasoning as a DAG at test time leads to more accurate, efficient multi-hop retrieval – and integrates seamlessly with RAG systems like Self-RAG.
📄 Paper: openreview.net/pdf?id=gi9aq...
3/3

openreview.net

Arno Solin @arnosolin.bsky.social · Apr 27

This work was born out of Prakhar's internship with Microsoft Research (\w Sukruta Prakash Midigeshi, Gaurav Sinha, Arno Solin, Nagarajan Natarajan, and Amit Sharma).
2/3

1

Arno Solin @arnosolin.bsky.social · Apr 27

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

1 7

Arno Solin @arnosolin.bsky.social · Apr 23

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

7

Arno Solin @arnosolin.bsky.social · Apr 23

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

1 1 5

Arno Solin @arnosolin.bsky.social · Apr 23

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

1 11

Arno Solin @arnosolin.bsky.social · Apr 21

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

2 3 10

Arno Solin @arnosolin.bsky.social · Apr 21

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

1 10

Arno Solin @arnosolin.bsky.social · Apr 21

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
📄 arxiv.org/abs/2408.02346

Exploiting Hankel-Toeplitz Structures for Fast Computation of Kernel Precision Matrices

The Hilbert-space Gaussian Process (HGP) approach offers a hyperparameter-independent basis function approximation for speeding up Gaussian Process (GP) inference by projecting the GP onto M basis fun...

arxiv.org

1

Arno Solin @arnosolin.bsky.social · Apr 21

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
📄 arxiv.org/abs/2405.17656

Alignment is Key for Applying Diffusion Models to Retrosynthesis

Retrosynthesis, the task of identifying precursors for a given molecule, can be naturally framed as a conditional graph generation task. Diffusion models are a particularly promising modelling approac...

arxiv.org

1 1 2

Arno Solin @arnosolin.bsky.social · Apr 21

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
📄 arxiv.org/abs/2411.18425

Streamlining Prediction in Bayesian Deep Learning

The rising interest in Bayesian deep learning (BDL) has led to a plethora of methods for estimating the posterior distribution. However, efficient computation of inferences, such as predictions, has b...

arxiv.org

1