arXiv subjects

Shouwei Gao

Publications and source records attributed to Shouwei Gao.

6 recordsLinked to original sources

arXiv2026

Evaluating LLM Coding Agents on SZ-Family Lossy Compression Across Architectures

Large language model (LLM) coding agents are increasingly applied to code translation and optimization, yet their effectiveness in performance-critical high-performance computing (HPC) settings remains poorly characterized. This paper evaluates LLM-based coding workflows on SZ-family error-bounded lossy compression kernels, which combine numerical constraints with memory-intensive and control-flow-heavy implementations. We study two representative CUDA workloads (SZp and SZx) and target two heterogeneous execution platforms: NVIDIA GPUs and Cerebras wafer-scale accelerators. Focusing on single-agent iterative generation, we analyze not only final throughput but also agent runtime behavior, including iteration patterns, sensitivity to prompt specification, and characteristic failure modes. Our results reveal a pronounced cross-architecture divergence. On GPUs, stronger models can achieve substantially higher throughput but exhibit increased sensitivity to prompt precision and optimization guidance, whereas on Cerebras the dominant challenge lies in producing runnable programs under a PE-centric spatial execution model. We further observe that LLM agents are more effective on modular kernels (SZx) than on tightly coupled bit-level pipelines (SZp), where structural dependencies hinder optimization progress. These findings suggest that evaluating LLM coding agents for HPC requires accounting for both performance outcomes and architecture-specific robustness, and that success on thread-based platforms does not directly transfer to spatial accelerators.

cs.DC↗

arXiv2026

LUMOS: Democratizing SciML Workflows with L0-Regularized Learning for Unified Feature and Parameter Adaptation

The rapid growth of scientific machine learning (SciML) has accelerated discovery across diverse domains, yet designing effective SciML models remains a challenging task. In practice, building such models often requires substantial prior knowledge and manual expertise, particularly in determining which input features to use and how large the model should be. We introduce LUMOS, an end-to-end framework based on L0-regularized learning that unifies feature selection and model pruning to democratize SciML model design. By employing semi-stochastic gating and reparameterization techniques, LUMOS dynamically selects informative features and prunes redundant parameters during training, reducing the reliance on manual tuning while maintaining predictive accuracy. We evaluate LUMOS across 13 diverse SciML workloads, including cosmology and molecular sciences, and demonstrate its effectiveness and generalizability. Experiments on 13 SciML models show that LUMOS achieves 71.45% parameter reduction and a 6.4x inference speedup on average. Furthermore, Distributed Data Parallel (DDP) training on up to eight GPUs confirms the scalability of

cs.LG↗

arXiv2026

FLYING SERVING: On-the-Fly Parallelism Switching for Large Language Model Serving

Production LLM serving must simultaneously deliver high throughput, low latency, and sufficient context capacity under non-stationary traffic and mixed request requirements. Data parallelism (DP) maximizes throughput by running independent replicas, while tensor parallelism (TP) reduces per-request latency and pools memory for long-context inference. However, existing serving stacks typically commit to a static parallelism configuration at deployment; adapting to bursts, priorities, or long-context requests is often disruptive and slow. We present Flying Serving, a vLLM-based system that enables online DP-TP switching without restarting engine workers. Flying Serving makes reconfiguration practical by virtualizing the state that would otherwise force data movement: (i) a zero-copy Model Weights Manager that exposes TP shard views on demand, (ii) a KV Cache Adaptor that preserves request KV state across DP/TP layouts, (iii) an eagerly initialized Communicator Pool to amortize collective setup, and (iv) a deadlock-free scheduler that coordinates safe transitions under execution skew. Across three popular LLMs and realistic serving scenarios, Flying Serving improves performance by up to $4.79\times$ under high load and $3.47\times$ under low load while supporting latency- and memory-driven requests.

cs.DC↗

arXiv2025

DOLMA: A Data Object Level Memory Disaggregation Framework for HPC Applications

Memory disaggregation is promising to scale memory capacity and improves utilization in HPC systems. However, the performance overhead of accessing remote memory poses a significant challenge, particularly for compute-intensive HPC applications where execution times are highly sensitive to data locality. In this work, we present DOLMA, a Data Object Level M emory dis Aggregation framework designed for HPC applications. DOLMA intelligently identifies and offloads data objects to remote memory, while providing quantitative analysis to decide a suitable local memory size. Furthermore, DOLMA leverages the predictable memory access patterns typical in HPC applications and enables remote memory prefetch via a dual-buffer design. By carefully balancing local and remote memory usage and maintaining multi-thread concurrency, DOLMA provides a flexible and efficient solution for leveraging disaggregated memory in HPC domains while minimally compromising application performance. Evaluating with eight HPC workloads and computational kernels, DOLMA limits performance degradation to less than 16% while reducing local memory usage by up to 63%, on average.

cs.DC↗

arXiv2024

PAC Learnability under Explanation-Preserving Graph Perturbations

Graphical models capture relations between entities in a wide range of applications including social networks, biology, and natural language processing, among others. Graph neural networks (GNN) are neural models that operate over graphs, enabling the model to leverage the complex relationships and dependencies in graph-structured data. A graph explanation is a subgraph which is an `almost sufficient' statistic of the input graph with respect to its classification label. Consequently, the classification label is invariant, with high probability, to perturbations of graph edges not belonging to its explanation subgraph. This work considers two methods for leveraging such perturbation invariances in the design and training of GNNs. First, explanation-assisted learning rules are considered. It is shown that the sample complexity of explanation-assisted learning can be arbitrarily smaller than explanation-agnostic learning. Next, explanation-assisted data augmentation is considered, where the training set is enlarged by artificially producing new training samples via perturbation of the non-explanation edges in the original training set. It is shown that such data augmentation methods may improve performance if the augmented data is in-distribution, however, it may also lead to worse sample complexity compared to explanation-agnostic learning rules if the augmented data is out-of-distribution. Extensive empirical evaluations are provided to verify the theoretical analysis.

cs.LG↗

arXiv2023

HurriCast: Synthetic Tropical Cyclone Track Generation for Hurricane Forecasting

The generation of synthetic tropical cyclone(TC) tracks for risk assessment is a critical application of preparedness for the impacts of climate change and disaster relief, particularly in North America. Insurance companies use these synthetic tracks to estimate the potential risks and financial impacts of future TCs. For governments and policymakers, understanding the potential impacts of TCs helps in developing effective emergency response strategies, updating building codes, and prioritizing investments in resilience and mitigation projects. In this study, many hypothetical but plausible TC scenarios are created based on historical TC data HURDAT2 (HURricane DATA 2nd generation). A hybrid methodology, combining the ARIMA and K-MEANS methods with Autoencoder, is employed to capture better historical TC behaviors and project future trajectories and intensities. It demonstrates an efficient and reliable in the field of climate modeling and risk assessment. By effectively capturing past hurricane patterns and providing detailed future projections, this approach not only validates the reliability of this method but also offers crucial insights for a range of applications, from disaster preparedness and emergency management to insurance risk analysis and policy formulation.

cs.LG↗