Simon Olsson
@smnlssn.bsky.social
Associate Professor at Chalmers. AI for molecular simulation and inverse design. WASP Fellow. ELLIS Member. https://userpage.fu-berlin.de/solsson/
😅😂 Thanks Shubhendu !
October 10, 2025 at 2:25 PM
😅😂 Thanks Shubhendu !
Excited to share this step toward scalable, transferable molecular simulation!
Lead by @viguera10.bsky.social
Thanks to wasp-sweden.org for funding and Chalmers for hosting us!
Link: arxiv.org/abs/2510.07589
Lead by @viguera10.bsky.social
Thanks to wasp-sweden.org for funding and Chalmers for hosting us!
Link: arxiv.org/abs/2510.07589
Transferable Generative Models Bridge Femtosecond to Nanosecond Time-Step Molecular Dynamics
Understanding molecular structure, dynamics, and reactivity requires bridging processes that occur across widely separated time scales. Conventional molecular dynamics simulations provide atomistic re...
arxiv.org
October 10, 2025 at 1:47 PM
Excited to share this step toward scalable, transferable molecular simulation!
Lead by @viguera10.bsky.social
Thanks to wasp-sweden.org for funding and Chalmers for hosting us!
Link: arxiv.org/abs/2510.07589
Lead by @viguera10.bsky.social
Thanks to wasp-sweden.org for funding and Chalmers for hosting us!
Link: arxiv.org/abs/2510.07589
Our model accelerates atomistic molecular dynamics by orders of magnitude while preserving physical accuracy, and generalizes to larger, unseen systems. This opens a route to exploring slow, functionally relevant molecular motions without sacrificing atomistic detail.
October 10, 2025 at 1:45 PM
Our model accelerates atomistic molecular dynamics by orders of magnitude while preserving physical accuracy, and generalizes to larger, unseen systems. This opens a route to exploring slow, functionally relevant molecular motions without sacrificing atomistic detail.
MGR recovers kinetic properties from umbrella-sampling with high fidelity
It also enables efficient Bayesian parameter inference for SDEs with sparse observations
Lead by Yan Wang when visiting us in Gothenburg, from Hao Wu's lab in Shanghai.
Preprint: arxiv.org/abs/2509.25872
It also enables efficient Bayesian parameter inference for SDEs with sparse observations
Lead by Yan Wang when visiting us in Gothenburg, from Hao Wu's lab in Shanghai.
Preprint: arxiv.org/abs/2509.25872
Marginal Girsanov Reweighting: Stable Variance Reduction via Neural Ratio Estimation
Recovering unbiased properties from biased or perturbed simulations is a central challenge in rare-event sampling. Classical Girsanov Reweighting (GR) offers a principled solution by yielding exact pa...
arxiv.org
October 3, 2025 at 10:12 AM
MGR recovers kinetic properties from umbrella-sampling with high fidelity
It also enables efficient Bayesian parameter inference for SDEs with sparse observations
Lead by Yan Wang when visiting us in Gothenburg, from Hao Wu's lab in Shanghai.
Preprint: arxiv.org/abs/2509.25872
It also enables efficient Bayesian parameter inference for SDEs with sparse observations
Lead by Yan Wang when visiting us in Gothenburg, from Hao Wu's lab in Shanghai.
Preprint: arxiv.org/abs/2509.25872
Standard Girsanov reweighting gives exact pathwise probability ratios, but its variance often explodes over long horizons, making it impractical for many applications. MGR deals with this problem through classifier-based density ratio estimation.
October 3, 2025 at 10:11 AM
Standard Girsanov reweighting gives exact pathwise probability ratios, but its variance often explodes over long horizons, making it impractical for many applications. MGR deals with this problem through classifier-based density ratio estimation.
we also found 20 or so over the weekend. Big batch of risotto for dinner tonight :)
September 21, 2025 at 6:18 PM
we also found 20 or so over the weekend. Big batch of risotto for dinner tonight :)
Congrats Martin! Fantastic news!
September 4, 2025 at 10:25 AM
Congrats Martin! Fantastic news!
Great news! Congrats @alextong.bsky.social and @aithyra.bsky.social !
September 3, 2025 at 3:19 PM
Great news! Congrats @alextong.bsky.social and @aithyra.bsky.social !