arXiv subjects

Sung Moon Ko

Publications and source records attributed to Sung Moon Ko.

11 recordsLinked to original sources

arXiv2025

Towards a Generalizable AI for Materials Discovery: Validation through Immersion Coolant Screening

Artificial intelligence (AI) has emerged as a powerful accelerator of materials discovery, yet most existing models remain problem-specific, requiring additional data collection and retraining for each new property. Here we introduce and validate GATE (Geometrically Aligned Transfer Encoder) -- a generalizable AI framework that jointly learns 34 physicochemical properties spanning thermal, electrical, mechanical, and optical domains. By aligning these properties within a shared geometric space, GATE captures cross-property correlations that reduce disjoint-property bias -- a key factor causing false positives in multi-criteria screening. To demonstrate its generalizable utility, GATE -- without any problem-specific model reconfiguration -- applied to the discovery of immersion cooling fluids for data centers, a stringent real-world challenge defined by the Open Compute Project (OCP). Screening billions of candidates, GATE identified 92,861 molecules as promising for practical deployment. Four were experimentally or literarily validated, showing strong agreement with wet-lab measurements and performance comparable to or exceeding a commercial coolant. These results establish GATE as a generalizable AI platform readily applicable across diverse materials discovery tasks.

cs.LG↗

arXiv2025

Geometric Embedding Alignment via Curvature Matching in Transfer Learning

Geometrical interpretations of deep learning models offer insightful perspectives into their underlying mathematical structures. In this work, we introduce a novel approach that leverages differential geometry, particularly concepts from Riemannian geometry, to integrate multiple models into a unified transfer learning framework. By aligning the Ricci curvature of latent space of individual models, we construct an interrelated architecture, namely Geometric Embedding Alignment via cuRvature matching in transfer learning (GEAR), which ensures comprehensive geometric representation across datapoints. This framework enables the effective aggregation of knowledge from diverse sources, thereby improving performance on target tasks. We evaluate our model on 23 molecular task pairs sourced from various domains and demonstrate significant performance gains over existing benchmark model under both random (14.4%) and scaffold (8.3%) data splits.

cs.LG↗

arXiv2024

Scalable Multi-Task Transfer Learning for Molecular Property Prediction

Molecules have a number of distinct properties whose importance and application vary. Often, in reality, labels for some properties are hard to achieve despite their practical importance. A common solution to such data scarcity is to use models of good generalization with transfer learning. This involves domain experts for designing source and target tasks whose features are shared. However, this approach has limitations: i). Difficulty in accurate design of source-target task pairs due to the large number of tasks, and ii). corresponding computational burden verifying many trials and errors of transfer learning design, thereby iii). constraining the potential of foundation modeling of multi-task molecular property prediction. We address the limitations of the manual design of transfer learning via data-driven bi-level optimization. The proposed method enables scalable multi-task transfer learning for molecular property prediction by automatically obtaining the optimal transfer ratios. Empirically, the proposed method improved the prediction performance of 40 molecular properties and accelerated training convergence.

cs.LG↗

arXiv2024

Task Addition in Multi-Task Learning by Geometrical Alignment

Training deep learning models on limited data while maintaining generalization is one of the fundamental challenges in molecular property prediction. One effective solution is transferring knowledge extracted from abundant datasets to those with scarce data. Recently, a novel algorithm called Geometrically Aligned Transfer Encoder (GATE) has been introduced, which uses soft parameter sharing by aligning the geometrical shapes of task-specific latent spaces. However, GATE faces limitations in scaling to multiple tasks due to computational costs. In this study, we propose a task addition approach for GATE to improve performance on target tasks with limited data while minimizing computational complexity. It is achieved through supervised multi-task pre-training on a large dataset, followed by the addition and training of task-specific modules for each target task. Our experiments demonstrate the superior performance of the task addition strategy for GATE over conventional multi-task methods, with comparable computational costs.

cs.LG↗

arXiv2024

Multitask Extension of Geometrically Aligned Transfer Encoder

Molecular datasets often suffer from a lack of data. It is well-known that gathering data is difficult due to the complexity of experimentation or simulation involved. Here, we leverage mutual information across different tasks in molecular data to address this issue. We extend an algorithm that utilizes the geometric characteristics of the encoding space, known as the Geometrically Aligned Transfer Encoder (GATE), to a multi-task setup. Thus, we connect multiple molecular tasks by aligning the curved coordinates onto locally flat coordinates, ensuring the flow of information from source tasks to support performance on target data.

cs.LG↗

arXiv2023

Geometrically Aligned Transfer Encoder for Inductive Transfer in Regression Tasks

Transfer learning is a crucial technique for handling a small amount of data that is potentially related to other abundant data. However, most of the existing methods are focused on classification tasks using images and language datasets. Therefore, in order to expand the transfer learning scheme to regression tasks, we propose a novel transfer technique based on differential geometry, namely the Geometrically Aligned Transfer Encoder (GATE). In this method, we interpret the latent vectors from the model to exist on a Riemannian curved manifold. We find a proper diffeomorphism between pairs of tasks to ensure that every arbitrary point maps to a locally flat coordinate in the overlapping region, allowing the transfer of knowledge from the source to the target data. This also serves as an effective regularizer for the model to behave in extrapolation regions. In this article, we demonstrate that GATE outperforms conventional methods and exhibits stable behavior in both the latent space and extrapolation regions for various molecular graph datasets.

cs.AI↗

arXiv2023

3D Denoisers are Good 2D Teachers: Molecular Pretraining via Denoising and Cross-Modal Distillation

Pretraining molecular representations from large unlabeled data is essential for molecular property prediction due to the high cost of obtaining ground-truth labels. While there exist various 2D graph-based molecular pretraining approaches, these methods struggle to show statistically significant gains in predictive performance. Recent work have thus instead proposed 3D conformer-based pretraining under the task of denoising, which led to promising results. During downstream finetuning, however, models trained with 3D conformers require accurate atom-coordinates of previously unseen molecules, which are computationally expensive to acquire at scale. In light of this limitation, we propose D&D, a self-supervised molecular representation learning framework that pretrains a 2D graph encoder by distilling representations from a 3D denoiser. With denoising followed by cross-modal knowledge distillation, our approach enjoys use of knowledge obtained from denoising as well as painless application to downstream tasks with no access to accurate conformers. Experiments on real-world molecular property prediction datasets show that the graph encoder trained via D&D can infer 3D information based on the 2D graph and shows superior performance and label-efficiency against other baselines.

cs.LG↗

arXiv2022

Grouping-matrix based Graph Pooling with Adaptive Number of Clusters

Graph pooling is a crucial operation for encoding hierarchical structures within graphs. Most existing graph pooling approaches formulate the problem as a node clustering task which effectively captures the graph topology. Conventional methods ask users to specify an appropriate number of clusters as a hyperparameter, then assume that all input graphs share the same number of clusters. In inductive settings where the number of clusters can vary, however, the model should be able to represent this variation in its pooling layers in order to learn suitable clusters. Thus we propose GMPool, a novel differentiable graph pooling architecture that automatically determines the appropriate number of clusters based on the input data. The main intuition involves a grouping matrix defined as a quadratic form of the pooling operator, which induces use of binary classification probabilities of pairwise combinations of nodes. GMPool obtains the pooling operator by first computing the grouping matrix, then decomposing it. Extensive evaluations on molecular property prediction tasks demonstrate that our method outperforms conventional methods.

cs.AI↗

arXiv2017

The rotation curve of a point particle in stringy gravity

Double Field Theory suggests to view the whole massless sector of closed strings as the gravitational unity. The fundamental symmetries therein, including the $\mathbf{O}(D,D)$ covariance, can determine unambiguously how the Standard Model as well as a relativistic point particle should couple to the closed string massless sector. The theory also refines the notion of singularity. We consider the most general, spherically symmetric, asymptotically flat, static vacuum solution to ${D=4}$ Double Field Theory, which contains three free parameters and consequently generalizes the Schwarzschild geometry. Analyzing the circular geodesic of a point particle in string frame, we obtain the orbital velocity as a function of $R/(M_{\scriptscriptstyle{\infty}}G)$ which is the dimensionless radial variable normalized by mass. The rotation curve generically features a maximum and thus non-Keplerian over a finite range, while becoming asymptotically Keplerian at infinity, $R/(M_{\scriptscriptstyle{\infty}}G)\rightarrow \infty$. The adoption of the string frame rather than Einstein frame is the consequence of the fundamental symmetry principle. Our result opens up a new scheme to solve the dark matter/energy problems by modifying General Relativity at `short' range of $R/(M_{\scriptscriptstyle{\infty}}G)$.

hep-th↗

arXiv2016

Search for a Higgs boson in the mass range from 145 to 1000 GeV decaying to a pair of W or Z bosons

A search for a heavy Higgs boson in the H to WW and H to ZZ decay channels is reported. The search is based upon proton-proton collision data samples corresponding to an integrated luminosity of up to 5.1 inverse femtobarns at sqrt(s) = 7 TeV and up to 19.7 inverse femtobarns at sqrt(s) = 8 TeV, recorded by the CMS experiment at the CERN LHC. Several final states of the H to WW and H to ZZ decays are analyzed. The combined upper limit at the 95% confidence level on the product of the cross section and branching fraction exclude a Higgs boson with standard model-like couplings and decays in the range 145 < m[H] < 1000 GeV. We also interpret the results in the context of an electroweak singlet extension of the standard model.

hep-ex↗

arXiv2012

Superconformal Yang-Mills quantum mechanics and Calogero model with OSp(N|2,R) symmetry

In spacetime dimension two, pure Yang-Mills possesses no physical degrees of freedom, and consequently it admits a supersymmetric extension to couple to an arbitrary number, N say, of Majorana-Weyl gauginos. This results in (N,0) super Yang-Mills. Further, its dimensional reduction to mechanics doubles the number of supersymmetries, from N to N+N, to include conformal supercharges, and leads to a superconformal Yang-Mills quantum mechanics with symmetry group OSp(N|2,R). We comment on its connection to AdS_2 \times S^{N-1} and reduction to a supersymmetric Calogero model.

hep-th↗