arXiv subjects

Yang Bao

Publications and source records attributed to Yang Bao.

3 recordsLinked to original sources

arXiv2026

NetRL: Powering Network Decisions via NetGRPO-Driven Reinforcement Learning

Network decision-making tasks, including adaptive bitrate streaming, load balancing, and congestion control, require policies to process heterogeneous states and actions while balancing near- and long-term consequences. Existing solutions generalize poorly across network tasks, require substantial task-specific redesign, and lack reliable evaluation of long- horizon action consequences. Methods that learn through online interaction additionally face unstable optimization. We present NetRL, a two-stage LLM-based reinforcement learning framework that combines supervised fine-tuning on demonstration trajectories with online policy optimization. A universal mixed-embedding interface combines textual task descriptions with precise numerical states, while a network action head produces valid decisions. At its core, NetGRPO evaluates candidate actions over multiple future trajectories, directs deeper search toward uncertain alternatives through confidence-aware budgeting, and performs group-relative, clipped, and KL-regularized updates. NetRL achieves 23.6% higher QoE than NetLLM, 12.2% lower job completion time than Genet, and 24.1% higher network utility than Orca, yielding state-of-the-art performance among the evaluated methods. Confidence-aware budgeting reduces end-to-end training time by 27.1% without reducing validation QoE. NetRL also converges stably and generalizes to unseen network traces without retraining. Code and model weights are available at the repository linked in this paper.

cs.NI↗

arXiv2026

Physics-Grounded Simulation Enables Zero-Shot Molecular and Reaction Analysis in Scanning Tunneling Microscopy

Quantitative identification of molecular products and reaction pathways from scanning tunneling microscopy images is difficult when labeled images are unavailable for a new chemical system. We develop a simulation-assisted framework for zero-shot analysis that requires only precursor information for a new target system, with no experimental images or annotations from that system used during model training. The framework couples two functional components: NanoSim generates physics-grounded synthetic images from predicted reaction products and sampled surface scenes, while NanoAnalysis translates molecular detections into product populations, spatial arrangements, and mechanistic interpretations. Applied to a representative on-surface reaction, the detector achieves a micro-averaged F1 score of 0.92 on experimental images. Analysis of the detected molecules reveals annealing-dependent population changes and reaction-site preferences consistent with experimental observations. Beyond this capability, its modular organization allows heterogeneous data from other systems to be used to improve each component independently, providing a scalable route for continued generalization.

q-bio.QM↗

arXiv2022

Predicting Corporate Risk by Jointly Modeling Company Networks and Dialogues in Earnings Conference Calls

Earnings conference calls are significant information events for volatility forecasting, which is essential for financial risk management and asset pricing. Although some recent volatility forecasting models have utilized the textual content of conference calls, the dialogue structures of conference calls and company relationships are almost ignored in extant literature. To bridge this gap, we propose a new model called Temporal Virtual Graph Neural Network (TVGNN) for volatility forecasting by jointly modeling conference call dialogues and company networks. Our model differs from existing models in several important ways. First, we propose to exploit more dialogue structures by encoding position, utterance, speaker role, and Q\&A segments. Second, we propose to encode the market states for volatility forecasting by extending the Gated Recurrent Units (GRU). Third, we propose a new method for constructing temporal company networks in which the messages can only flow from temporally preceding to successive nodes, and extend the Graph Attention Networks (GAT) for modeling company relationships. We collect conference call transcripts of S\&P500 companies from 2008 to 2019, and construct a dataset of conference call dialogues with additional information on dialogue structures and company networks. Empirical results on our dataset demonstrate the superiority of our model over competitive baselines for volatility forecasting. We also conduct supplementary analyses to examine the effectiveness of our model's key components and interpretability.

cs.CL↗