Date/Time : 15 Jul 2026 16:00
Venue : S17 #04-04 (Seminar Room 3)
Speaker : Wang Hao, Sichuan University

Solid-state phase transitions typically follow a Minimum Energy Path (MEP). Classical methods such as G-SSNEB compute the MEP in the full high-dimensional configuration space, which becomes prohibitive when paths must be scanned over pressures, defect types, and system sizes. We propose the Preset Structure Strain Path (PSSP) method, which uses feature extraction and transfer learning to compute MEPs efficiently across these conditions. PSSP is based on the fact that in defective systems, lattice strain sets the direction of the MEP, while the local atomistic environment changes little but governs the energy. PSSP fully exploits this separation by presetting the displacement field as a transferable feature and optimizing only the lattice evolution in a low-dimensional space. By doing this, the strain paths learned in one setting can be reused under different pressures, defects and system sizes. The advantage of PSSP is verified that it reduces the energy and force evaluations to less than 2% of G-SSNEB but preserving the accuracy of the energy barrier in a satisfactory level.

Add to calendar :

Date/Time : 03 Jul 2026 10:00
Venue : S17 #04-05 (Seminar Room 2)
Speaker : Laura Schaposnik, University of Illinois Urbana-Champaign

The moduli space of Higgs bundles carries several natural geometric structures, making it a central example in the study of mirror symmetry and gauge theory. In this lecture we will explore special subspaces of Higgs bundle moduli spaces, often called branes, and explain how they arise from symmetries, real forms, and geometric constraints. We will discuss their interpretation through flat connections and surface group representations, and describe how different types of branes are related through natural correspondences, and how they can be seen through the Hitchin fibration. The goal is to illustrate how Higgs bundles connect algebraic geometry, representation theory, and ideas from mathematical physics.

Add to calendar :

Date/Time : 29 Jun 2026 10:00
Venue : S17 #04-05 (Seminar Room 2)
Speaker : Laura Schaposnik, University of Illinois Urbana-Champaign

Higgs bundles provide a rich meeting point between algebraic geometry, differential geometry, gauge theory, and mathematical physics. In this lecture we will introduce the basic objects: vector bundles on a Riemann surface equipped with Higgs fields, their stability condition, and the construction of the corresponding moduli spaces. We will then discuss the Hitchin fibration, one of the central structures in the theory, and explain how spectral curves encode much of the geometry of Higgs bundles. The lecture will emphasize the interplay between moduli spaces, integrable systems, and the geometric structures that make Higgs bundles such powerful objects.

Add to calendar :

Date/Time : 26 Jun 2026 13:00
Venue : via Zoom
Speaker : Yang Liu, National University of Singapore

Join Zoom Meeting
https://nus-sg.zoom.us/j/89586663814?pwd=JXmq5ImFrfOM72MJpf2lyririCPiQs.1

Meeting ID: 895 8666 3814
Passcode: 942739

We introduce Evolutionary Ensemble (EvE), a decentralized framework that organizes existing, highly capable coding agents into a live, co-evolving system for math proof, algorithm discovery, and beyond. By maintaining two co-evolving populations, namely downstream task solvers and agent guidance states, the system keeps evolving the agents based on the marginal gains they contribute to the solvers. Relying solely on a $200 Codex subscription without additional computational resources or human intervention, EvE was able to solve a mathematical problem in just one day that previously took one of Zhang Lei’s PhD students an entire year to complete. We will also show some results of applying EvE to algorithm discovery, control design, and beyond. We will discuss how to accelerate our research with EvE.

EvE is open-source and actively maintained: https://github.com/scaling-group/eve

Add to calendar :

Date/Time : 24 Jun 2026 11:00
Venue : S17 #06-11 (Seminar Room 6)
Speaker : Cheng Xiuyuan, Duke University

Wasserstein regularization has become a common tool in learning problems over distributions, while its practical computation remains challenging, especially in high dimensions. This talk presents a framework for Wasserstein-regularized learning based on directly parameterizing transport maps with neural networks. Rather than working with dual potentials or entropic relaxations, we explicitly model the transport map, which enables principled optimization in Wasserstein space as well as scalable out-of-sample evaluation. One example is the implementation of Wasserstein-2 proximal steps via flow networks that realize the Jordan–Kinderlehrer–Otto (JKO) scheme, yielding a variational interpretation of flow-based generative models. The talk will also discuss a minimax regime motivated by distributionally robust optimization (DRO), where transport is chosen adversarially. In this setting, neural transport maps provide a practical framework for studying Wasserstein minimax problems and representing worst-case distributions. Together, these examples highlight the flexibility of transport-map formulations for incorporating Wasserstein geometry into learning problems over distributions.

Add to calendar :

Date/Time : 17 Jun 2026 15:00
Venue : S17 #04-04 (Seminar Room 3)
Speaker : Guo Weihong, Case Western Reserve University

Variational model based and deep learning data-based approaches are distinct tools for image analysis and computer vision, each with its own strengths and limitations. Variational models minimize energy functionals which encode desired properties like smoothness, boundary sharpness, etc. The models provide a clear interpretation and control over image features by picking one or multiple regularization terms. Instead of relying on handcrafted features, deep learning models learn image features from data during the training process. These models can directly map input images to desired outputs, making the process highly automated and efficient. DL methods lack interpretability in terms of how exactly the model is making its decisions. Without enough data or proper regularization, models may overfit to training data and the results miss regularity. In this talk, we present our recent research done in fusing the strength of variational models and deep learning models for image segmentation methods that have better mathematical interpretability, less overfitting and better regularity.

Add to calendar :

Date/Time : 12 Jun 2026 16:00
Venue : S17 #04-05 (Seminar Room 2)
Speaker : Shuai Cheng Li, City University of Hong Kong

Breakage-fusion-bridge (BFB) is a complex rearrangement that leads to tumor malignancy. Existing models for detecting BFBs rely on the ideal BFB hypothesis, ruling out the possibility of BFBs entangled with other structural variations, that is, complex BFBs. We propose an algorithm, Ambigram, to identify complex BFB and reconstruct the rearranged structure of the local genome during the cancer subclone evolution process. Ambigram handles data from short, linked, long, and single-cell sequences, and optical mapping technologies. Ambigram successfully deciphers the gold- or silver-standard complex BFBs against the state-of-the-art in multiple cancers. Ambigram dissects the intratumor heterogeneity of complex BFB events with single-cell reads from melanoma and gastric cancer. Furthermore, applying Ambigram to liver and cervical cancer data suggests that the BFB mechanism may mediate oncovirus integrations. BFB also exists in noncancer genomics.
Investigating the complete human genome reference with Ambigram suggests that the BFB mechanism may be involved in two genome reorganizations of Homo Sapiens during evolution. Moreover, Ambigram discovers the signals of recurrent foldback inversions and complex BFBs in whole genome data from the 1000 Genome Project, and congenital heart diseases, respectively.

Add to calendar :

Date/Time : 10 Jun 2026 15:00
Venue : S17 #04-04 (Seminar Room 3)
Speaker : Wilderich Tuschmann, Karlsruher Institute of Technology

At its core, the Manifold Hypothesis asserts that real-world, high-dimensional data is not uniformly or randomly distributed throughout its high-dimensional “ambient” space, but concentrated on or near a low-dimensional manifold (or a collection of manifolds) embedded within that high-dimensional ambient space.

In my talk, I will discuss reasons and facts that speak for as well as against this hypothesis as well as address geometric alternatives.

Add to calendar :

Date/Time : 29 May 2026 16:00
Venue : S17 #04-05 (Seminar Room 2)
Speaker : Yufeng Wu, University of Connecticut

Single cell sequencing technologies are generating DNA sequences for individual cells of an organism. Such data enables novel study of single cell evolution. Cell lineage tree is an evolutionary model for single cells, where the leaves of the tree are the extant cells and the internal nodes are the progenitor cells. Reconstructing cell lineage tree from single cell data has been actively studied recently. The main computational challenges for cell lineage tree inference are the inherent noises in single cell data and also the size of the data. In this talk, I will first introduce perfect phylogeny, a simple evolutionary tree model that has been used extensively in cancer and single cell evolution. Then I will present my own work ScisTree, a probabilistic cell lineage tree inference method based on the perfect phylogeny model. Then, I will focus on presenting ScisTree2, a state-of-the-art cell lineage tree inference method that is orders of magnitude faster than ScisTree. I will show some results of ScisTree2 on simulated and biological data.

Add to calendar :