
Sahel Iqbal
I’m Sahel Iqbal, a third-year PhD student at Aalto University, Finland, where I work with Simo Särkkä. My main research focus is on developing accurate and efficient Monte Carlo algorithms for reinforcement learning and Bayesian experimental design (BED). Recently, I have also developed an interest in how similar algorithms can be used for inference-time alignment of diffusion and large language models.
For academic details, see my resume and my Google Scholar profile. I can be contacted on X and at my email sahel[dot]iqbal[at]aalto[dot]fi.
Outside work, my time is mostly taken up by reading, lifting weights, and writing JAX code. The projects that I’m actively working on are available on GitHub.
Recent News
- 2025-09: Our paper Sequential Monte Carlo for Policy Optimization in Continuous POMDPs has been accepted to NeurIPS 2025!
- 2025-07: If you’re attending MCM 2025, my coauthor Adrien Corenflos will be giving a talk on our joint work on BED.
- 2025-06: Gave a talk on using particle filters for amortized BED at the Accelerating statistical inference and experimental design with machine learning workshop at the Isaac Newton Institute for Mathematical Sciences.
- 2024-12: Presented a poster at the Bayesian Decision-making and Uncertainty workshop at NeurIPS 2024 in Vancouver.
Posts
2025-10-09
Steering Language Models with Sequential Monte Carlo — How to give your language model the blues with SMC.
2025-09-24
Using LaTeX Snippets in Markdown Files in Neovim — How I adapted my LuaSnip LaTeX snippets to work in Markdown for easier math note-taking in Neovim.
2025-06-08
Expected Information as Expected Utility — I discuss the article of the same name by José M. Bernardo from 1979, which shows that the expected information gain, a popular metric used in Bayesian experimental design, is itself a solution to a maximum expected utility problem under some assumptions on the utility function.
Featured Publications
* denotes equal contribution.
Hany Abdulsamad*, Sahel Iqbal*, Simo Särkkä (2025). Sequential Monte Carlo for policy optimization in continuous POMDPs. NeurIPS. arXiv. Code.
Mahdi Nasiri, Sahel Iqbal, Simo Särkkä (2025). Physics-informed machine learning for grade prediction in froth flotation. Minerals Engineering. Link.
Sahel Iqbal, Hany Abdulsamad, Sara Pérez-Vieites, Simo Särkkä, Adrien Corenflos (2024). Recursive nested filtering for efficient amortized Bayesian experimental design. NeurIPS workshop on Bayesian Decision-making and Uncertainty. arXiv. Code.
Sahel Iqbal, Hany Abdulsamad, Tripp Cator, Ulisses Braga-Neto, Simo Särkkä (2024). Parallel-in-time probabilistic solutions for time-dependent nonlinear partial differential equations. IEEE MLSP. Link. Code.
Sahel Iqbal, Adrien Corenflos, Simo Särkkä, Hany Abdulsamad (2024). Nesting particle filters for experimental design in dynamical systems. ICML. arXiv. Code.