1 citations · 1 across the 10 of their papers we have counts for
14 papers
Solving Without Stopping: On-Policy Distillation at Small Scale
Hongyang Li, Yiming Zhu, Xiao Li +3
On-policy distillation, where a student learns from a stronger teacher's feedback on its own outputs, is a common way to pass reasoning to smaller models. We analyze what it transf…
Unlocking the Critic: Reward-Free Policy Optimization for LLM Post-Training
Hongyang Li, Xiao Li, Caesar Wu +3
Recent approaches to reinforcement learning (RL) post-training for large language models increasingly remove the critic to reduce training instability and memory overhead. Even whe…
What Does Post-Training Change in Multilingual Reasoning?
Hongyang Li, Xiao Li, Caesar Wu +2
Open-source reasoning models provide unequal access to reasoning capability across languages. When a model can solve a problem but cannot deliver a complete solution in the user's…
Beyond Compression: Diagnosing How Post-Training Changes Mathematical Reasoning
Hongyang Li, Yiming Zhu, Xiao Li +3
Post-training is central to mathematical reasoning in modern large language models (LLMs), but endpoint pass@1 alone underidentifies what has changed. Gains may reflect newly reach…
Reward-Guided Autoregressive Graph Generation for Efficient Multi-Agent Communication Topology Design
Poomphob Suwannapichat, Boonyarit Changaival, Caesar Wu +1
LLM-based Multi-Agent Systems (MAS) achieve strong performance on complex reasoning tasks by coordinating multiple agents, but at the cost of substantial token consumption. Recent…
Trustworthy Efficient Communication for Distributed Learning using LQ-SGD Algorithm
Hongyang Li, Lincen Bai, Caesar Wu +3
We propose LQ-SGD (Low-Rank Quantized Stochastic Gradient Descent), an efficient communication gradient compression algorithm designed for distributed training. LQ-SGD further deve…
Lightweight Trustworthy Distributed Clustering
Hongyang Li, Caesar Wu, Mohammed Chadli +2
Ensuring data trustworthiness within individual edge nodes while facilitating collaborative data processing poses a critical challenge in edge computing systems (ECS), particularly…
A Unified Hyperparameter Optimization Pipeline for Transformer-Based Time Series Forecasting Models
Jingjing Xu, Caesar Wu, Yuan-Fang Li +2
Transformer-based models for time series forecasting (TSF) have attracted significant attention in recent years due to their effectiveness and versatility. However, these models of…
Trustworthiness of Stochastic Gradient Descent in Distributed Learning
Hongyang Li, Caesar Wu, Mohammed Chadli +2
Distributed learning (DL) uses multiple nodes to accelerate training, enabling efficient optimization of large-scale models. Stochastic Gradient Descent (SGD), a key optimization a…
Dataset | Mindset = Explainable AI | Interpretable AI
Caesar Wu, Rajkumar Buyya, Yuan Fang Li +1
We often use "explainable" Artificial Intelligence (XAI)" and "interpretable AI (IAI)" interchangeably when we apply various XAI tools for a given dataset to explain the reasons th…
Heterogeneity: An Open Challenge for Federated On-board Machine Learning
Maria Hartmann, Grégoire Danoy, Pascal Bouvry
The design of satellite missions is currently undergoing a paradigm shift from the historical approach of individualised monolithic satellites towards distributed mission configura…
Survey and Taxonomy: The Role of Data-Centric AI in Transformer-Based Time Series Forecasting
Jingjing Xu, Caesar Wu, Yuan-Fang Li +2
Alongside the continuous process of improving AI performance through the development of more sophisticated models, researchers have also focused their attention to the emerging con…
Training Green AI Models Using Elite Samples
Mohammed Alswaitti, Roberto Verdecchia, Grégoire Danoy +2
The substantial increase in AI model training has considerable environmental implications, mandating more energy-efficient and sustainable AI practices. On the one hand, data-centr…
Transformer Multivariate Forecasting: Less is More?
Jingjing Xu, Caesar Wu, Yuan-Fang Li +1
In the domain of multivariate forecasting, transformer models stand out as powerful apparatus, displaying exceptional capabilities in handling messy datasets from real-world contex…