Sign in

Isaac Huidobro Meezs

@ishume.bsky.social
58 followers 116 following 24 posts

🇲🇽 in 🇨🇦. Theoretical chemist. Interested in Quantum (Non-adiabatic) Dynamics, Quantum Computing, Machine Learning and Electronic Structure.

PostsRepliesMedia
Isaac Huidobro Meezs @ishume.bsky.social · 08/10/2026
New preprint! We use binary optimization to group commuting Pauli operators, reducing sampling costs by 25% on molecular benchmarks relative to ICS. The found groupings are transferable to nearby geometries, and the method still performs well on lattice Hamiltonians :) 📄 arxiv.org/abs/2610.10339
arxiv.org
Binary Optimization of Measurement Groupings for Quantum Energy Estimation
Repeated measurements can dominate the resources required for quantum energy estimation, making the choice of which Pauli observables to measure together a central optimization problem for variational...
010 0
Isaac Huidobro Meezs @ishume.bsky.social · 10/09/2026
Excited to share Grinch! We use GFlowNets to build graphs of quantum optical experiments, guiding design with logic-based masking functions and finding multiple high-fidelity solutions for target states and photonic gates while respecting connectivity constraints :) arxiv.org/abs/2609.08073
arxiv.org
Bottom-Up Design of Quantum Optical Experiments Using Discrete Generative Models
Designing quantum optical experiments requires searching over discrete circuit topologies and continuous parameters, often with multiple realizations of the same target state. Graph-based methods comm...
020 0
Isaac Huidobro Meezs @ishume.bsky.social · 08/07/2026
My view is to start with cheap classical/ML covariances for groupings, update them as early measurement data accumulates (not necessarily at every VQE step, but when the state difference across VQE steps justifies it), and update the groupings to improve the measurement strategy.
010 0
Isaac Huidobro Meezs @ishume.bsky.social · 08/07/2026
Across the full benchmark, SI-ICS was better than all VarSI-ICS variants in only 5/418 cases, and all used approximate CISD covariances. So, covariance mismatch can hurt, but it was not the dominant behavior in our tests. Of course, the closer you are to the wfn variances on the device, the better
110 0
Isaac Huidobro Meezs @ishume.bsky.social · 08/07/2026
Great question! In this paper, we tested this mainly by using CISD covariances as an approximate proxy for the state and then evaluating the final measurement cost using exact ground-state variances. In that setting, the VarSI/ICS initializations were fairly robust.
110 0
Isaac Huidobro Meezs @ishume.bsky.social · 07/07/2026
Across 130 molecular Hamiltonians, VarSI initializations reduce ICS measurement costs by up to 70% relative to SI, with average reductions around 10–15%. These savings can translate into about an hour less QPU runtime for tasks like building a PES 😬
000 0
Isaac Huidobro Meezs @ishume.bsky.social · 07/07/2026
Overlapping grouping methods are among the strongest approaches for reducing measurement cost, but they still rely on a non-overlapping initialization, typically from Sorted Insertion (SI). VarSI uses the same covariance information these methods already require to build better starting groups.
100 0
Isaac Huidobro Meezs @ishume.bsky.social · 07/07/2026
Now on arXiv :) arxiv.org/abs/2607.02794 Measurement cost is a major bottleneck in variational quantum algorithms. We introduce VarSI: a heuristic that reduces Hamiltonian measurement costs and improves overlapping-method initializations, thereby cutting measurement counts and QPU runtime.
arxiv.org
Reducing quantum measurements in qubit-based overlapping grouping methods for quantum energy estimation through better initializations
The measurement cost for estimating expectation values of Hamiltonians is a central bottleneck in variational quantum algorithms. Grouping strategies significantly reduce this cost, with overlapping t...
100 0
Isaac Huidobro Meezs @ishume.bsky.social · 06/07/2026
This highlights how we can incorporate hardware-driven rewards to optimize other metrics in quantum computing using GFlowNets. Our findings also indicate that improving the initialization of overlapping grouping methods offers an opportunity to further reduce measurement costs (Work on this soon!)
000 0
Isaac Huidobro Meezs @ishume.bsky.social · 06/07/2026
An update on our work on measurement reduction in quantum computing for chemistry! We use our framework to reduce both measurement and two-qubit gate costs through reward-driven sampling. We also show that our groupings are better initializations for overlapping methods arxiv.org/abs/2509.154...
130 0
Isaac Huidobro Meezs @ishume.bsky.social · 06/07/2026
Really happy to have participated in the second AlgoLab Summer School at Université de Sherbrooke. A great couple of weeks filled with excellent lectures, stimulating discussions, and new ideas to explore :)
000 0
Isaac Huidobro Meezs @ishume.bsky.social · 06/07/2026
Had a great time at the Canadian Societies for Chemistry and Chemical Engineering Conference x2026 in Toronto! It was a pleasure to attend with the ChemAI lab and talk about our work on measurement optimization for quantum computing in chemistry, and learn from many inspiring talks.
000 0
Reposted by Isaac Huidobro Meezs
Zoe Holmes @qzoeholmes.bsky.social · 19/12/2025
🥈In second place we then had a thought provoking analysis on the measurement problem… are they correct to make the fire quantum mechanics and the child the measurement problem or should it be the other way around? 🤔
162 0
Isaac Huidobro Meezs @ishume.bsky.social · 21/11/2025
Honored to receive SandboxAQ’s Research Excellence Award! Our project introduces generative derandomized classical shadows to reduce measurement costs by learning better unitary sampling. Excited to share results soon! tinyurl.com/hkxamfz4
tinyurl.com
"Quantum measurements efficiency without accuracy loss: Isaac Huidobro Meezs' research" | SandboxAQ posted on the topic | LinkedIn
“How can we make quantum measurements more efficient without sacrificing accuracy?” Isaac Huidobro Meezs is diving into this research question at McMaster University in Canada. Isaac was drawn to hi...
040 0
Reposted by Isaac Huidobro Meezs
Rafael Wagner @quantumrw.bsky.social · 21/11/2025
arxiv.org/abs/2511.15806 If correct, that's a huge result!
arxiv.org
031 0
Reposted by Isaac Huidobro Meezs
Physical Chemistry Chemical Physics @pccp.rsc.org · 20/11/2025
Our latest themed collection ‘Festschrift for Christel Marian: The role of electron spin in molecular photochemical and photophysical processes: Theory and experiment’ is now online and free to access until 14th February 2026! 🧪 #chemsky Read the collection: pubs.rsc.org/en/journals/...
An R-S-C promotional graphic for the P-C-C-P Festschrift for Christel Marian: The role of electron spin in molecular photochemical and photophysical processes themed collection. The graphic includes photos of the guest editors, Shirin Faraj, Andres Dreuw, Bernd Engels, Martin Kleinschmidt. The tagline reads “Free to access”.
032 0
Isaac Huidobro Meezs @ishume.bsky.social · 28/10/2025
A bit late in sharing this, but truly honored to share our progress on flow-based models for quantum algorithms and quantum dynamics at my home university, UNAM, for the 60th anniversary of the Faculty of Chemistry’s Graduate School. Grateful for the invitation and insightful discussions!
050 0
Isaac Huidobro Meezs @ishume.bsky.social · 22/09/2025
Really happy to see this one out! We are currently working on including recent developments in GFlowNets to enhance our protocol and explore other constraints relevant to current quantum devices, like gate fidelity and circuit implementation restrictions.
020 0
Reposted by Isaac Huidobro Meezs
Rodrigo A. Vargas-Hdz @rovargash.bsky.social · 11/08/2025
Exciting news! 🎉 My student, @ishume.bsky.social , has been selected as one of the 2025 @sandboxaq.bsky.social Scholarship Recipients! We can’t wait to show you what GenAI can do for quantum computing, stay tuned!! www.sandboxaq.com/company/scho... @mcmasteruniversity.bsky.social
sandboxaq.com
SandboxAQ Scholarship Program
SandboxAQ is proud to announce the launch of two exciting initiatives designed to support PhD students pursuing cutting-edge research in AQ (AI and Quantum) technology.
051 0
Reposted by Isaac Huidobro Meezs
Sam Blau @samblau.bsky.social · 14/05/2025
The Open Molecules 2025 dataset is out! With >100M gold-standard ωB97M-V/def2-TZVPD calcs of biomolecules, electrolytes, metal complexes, and small molecules, OMol is by far the largest, most diverse, and highest quality molecular DFT dataset for training MLIPs ever made 1/N
54610 0
Reposted by Isaac Huidobro Meezs
Alán Aspuru-Guzik @aspuru.bsky.social · 06/05/2025
A very important from our group #compchemsky #aiforscience #quantumchemistry El Agente calculates complex quantum chemistry calculations from natural language! Please re-share with your network. @thematterlab.bsky.social @uoft.bsky.social #nvidia
03010 0
Reposted by Isaac Huidobro Meezs
Chemical Science @chemicalscience.rsc.org · 08/03/2025
Anna Painelli’s research focuses on molecular spectroscopy. Her recent Perspective in Chemical Science seeks to shed light on thermally-activated delayed fluorescence, which has a variety of applications from organic LEDs to metal-free photocatalysis. buff.ly/p1t6aF4 #ChemSky
doi.org
Shedding light on thermally-activated delayed fluorescence
Thermally activated delayed fluorescence (TADF) is a hot research topic in view of its impressive applications in a wide variety of fields from organic LEDs to photodynamic therapy and metal-free…
1113 0
Reposted by Isaac Huidobro Meezs
Guillermo del Toro @realgdt.bsky.social · 03/03/2025
Hurra!!!!!
342661246 0
Reposted by Isaac Huidobro Meezs
Rachel Crespo Otero @rcrespootero.bsky.social · 20/12/2024
I have a PhD scholarship (UCL-A*STAR) in "Data-driven analysis of excited states in metal-organic frameworks". This is a collaboration with @mbsullivan.bsky.social. See the details of the scholarship following this link: www.findaphd.com/phds/project...
findaphd.com
A 3-Year PhD Studentship in Data-driven analysis of excited states in metal-organic frameworks at University College London on FindAPhD.com
PhD Project - A 3-Year PhD Studentship in Data-driven analysis of excited states in metal-organic frameworks at University College London, listed on FindAPhD.com
2106 0
Reposted by Isaac Huidobro Meezs
David Abergel @davidabergel.bsky.social · 26/11/2024
⚛️🧪 If you follow condensed-matter physics, you may have noticed the theoretical idea of the quantum geometric tensor of a crystalline material has been discussed a lot recently. Now, Riccardo Comin's group have shown how to measure it using a fancy ARPES technique. Free-to-read link: rdcu.be/d1wES
rdcu.be
Measurements of the quantum geometric tensor in solids
Nature Physics - Experiments that directly probe the quantum geometric tensor in solids have not been reported. Now, the quantum metric and spin Berry curvature—dual components of the quantum...
0153 0
Reposted by Isaac Huidobro Meezs
Mario Barbatti @mbarbatti.bsky.social · 17/12/2024
Dynamics simulations of light-excited molecules are still a big challenge. If we need long timescales, comput. costs explode, and the methods available may not even work well. Here are the results of 4y testing the main methods: They're expensive, but they all do a good job in the predictions.🧪
doi.org
Assessing Nonadiabatic Dynamics Methods in Long Timescales
Nonadiabatic dynamics simulations complement time-resolved experiments by revealing ultrafast excited-state mechanistic information in photochemical reactions. Understanding the relaxation mechanisms ...
1194 0
Isaac Huidobro Meezs @ishume.bsky.social · 15/12/2024
Great week at @neuripsconf.bsky.social ! Excited to present my work on GFlowNets for Chemistry in Quantum Computers at the Machine Learning and the Physical Sciences workshop with @rovargash.bsky.social 😬 #NeurIPS2024 #ml4ps2024 #ml4phys #compchem
031 0
Isaac Huidobro Meezs @ishume.bsky.social · 10/12/2024
Great tutorials today!
010 0
Isaac Huidobro Meezs @ishume.bsky.social · 10/12/2024
Stop by to chat at poster 167 about our GFlowNets approach for Hamiltonian decomposition, from @rovargash.bsky.social's lab :) arxiv.org/abs/2410.16041 #compchem
arxiv.org
GFlowNets for Hamiltonian decomposition in groups of compatible operators
Quantum computing presents a promising alternative for the direct simulation of quantum systems with the potential to explore chemical problems beyond the capabilities of classical methods. However, c...
020 0
Reposted by Isaac Huidobro Meezs
Eli LF @elilevensonfalk.bsky.social · 03/03/2024
Maybe my favorite so far
141 0
Reposted by Isaac Huidobro Meezs
arxiv quant-ph @arxiv-quant-ph.bsky.social · 02/12/2024
Isaac Layton, Jonathan Oppenheim The classical-quantum limit arxiv.org/abs/2310.18271
011 0
Reposted by Isaac Huidobro Meezs
Mathurin Massias @mathurinmassias.bsky.social · 27/11/2024
Anne Gagneux, Ségolène Martin, @quentinbertrand.bsky.social Remi Emonet and I wrote a tutorial blog post on flow matching: dl.heeere.com/conditional-... with lots of illustrations and intuition! We got this idea after their cool work on improving Plug and Play with FM: arxiv.org/abs/2410.02423
1235299 0
Reposted by Isaac Huidobro Meezs
Daumann Lab @daumannl.bsky.social · 28/11/2024
Hi #ChemSky! Please enjoy this video from my first year #chemistry lecture demonstrating the principle of triboluminescence by crushing crystals of the copper complex, [Cu(NCS)(py)2(PPh3)]! 😍 This also works with simple quartz crystals #RealTimeChem 🧪🥽
613032 0
Isaac Huidobro Meezs @ishume.bsky.social · 20/11/2024
Here's an infographic on our most recent work on GFlowNets for Hamiltonian decomposition in groups of compatible operators, aimed for a broader audience ⚛️ #compchemsky #chemsky #quantumcomputing #compchem
030 0
Isaac Huidobro Meezs @ishume.bsky.social · 20/11/2024
We are currently working on improving our approach by introducing faster credit assignment and including better measurement estimates. We will also extend it to include other grouping techniques.
000 0
Isaac Huidobro Meezs @ishume.bsky.social · 20/11/2024
We consider the coloring problem of the complementary commutativity graph. GFlowNets learns a policy to sample grouping configurations for molecular Hamiltonians. The probability to assign a color to each term is updated according to a measurement based reward.
100 0
Isaac Huidobro Meezs @ishume.bsky.social · 20/11/2024
Our most recent work, accepted to ML4Phys workshop at NeurIPS! 📝: arxiv.org/abs/2410.16041 We employed GFlowNets to generate groupings of molecular Hamiltonians aimed to reduce the required number of measurements. 🧑‍💻 #compchem #quantumcomputing #chemsky
150 0
Isaac Huidobro Meezs @ishume.bsky.social · 20/11/2024
A couple of works which inspired my interest on non-adiabatic dynamics! In both I employed non-adiabatic transition state theory and were able to explain the magnetic moment of the Pt13[1] and Al6[2] clusters via spin-transitions ⚛️ [1] pubs.acs.org/doi/10.1021/... [2] pubs.acs.org/doi/10.1021/...
pubs.acs.org
Effect of Tunneling-Assisted Intersystem Crossing on the Magnetic Behavior of the Pt13 Cluster
We show the existence of a dynamic spin equilibrium introduced by intersystem crossing (ISC) between several spin multiplicity states in the Pt13 cluster. Employing weak-coupling transition probabilit...
000 0
Isaac Huidobro Meezs @ishume.bsky.social · 20/11/2024
In this work, we studied haptotropic shifts on transition-metal–[5]cumulene complexes employing QTAIM and ELI-D. We showed that the M–C bond closer to terminal groups is significantly weakened by steric interactions, driving equilibrium. chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/...
chemistry-europe.onlinelibrary.wiley.com
The Role of Bulkiness in Haptotropic Shifts of Metal–Cumulene Complexes
To correctly predict the direction of haptotropic shifts in metal–cumulene complexes, not only is the bulkiness of terminal groups important, but also the orientation of these groups with respect to ...
000 0