18 citations · 47 across the 27 of their papers we have counts for
12 papers · 1 filter
Rethinking Groups in Critic-Free RLVR
Yihong Wu, Liheng Ma, Lingfeng Xiao +4
Reinforcement learning (RL) has become a central paradigm for post-training large language models. Existing critic-free RL methods typically generate a group of rollouts for the sa…
TrasMuon: Trust-Region Adaptive Scaling for Orthogonalized Momentum Optimizers
Peng Cheng, Jiucheng Zang, Qingnan Li +6
Muon-style optimizers leverage Newton-Schulz (NS) iterations to orthogonalize updates, yielding update geometries that often outperform Adam-series methods. However, this orthogona…
It Takes Two: Your GRPO Is Secretly DPO
Yihong Wu, Liheng Ma, Lei Ding +9
GRPO has emerged as a prominent reinforcement learning algorithm for post-training LLMs. Unlike critic-based methods, GRPO computes advantages by estimating the \emph{value baselin…
Omni-Thinker: Scaling Multi-Task RL in LLMs with Hybrid Reward and Task Scheduling
Derek Li, Jiaming Zhou, Leo Maxime Brunswic +8
The pursuit of general-purpose artificial intelligence depends on large language models (LLMs) that can handle both structured reasoning and open-ended generation. We present Omni-…
SKOLR: Structured Koopman Operator Linear RNN for Time-Series Forecasting
Yitian Zhang, Liheng Ma, Antonios Valkanas +2
Koopman operator theory provides a framework for nonlinear dynamical system analysis and time-series forecasting by mapping dynamics to a space of real-valued measurement functions…
Sparse Decomposition of Graph Neural Networks
Yaochen Hu, Mai Zeng, Ge Zhang +4
Graph Neural Networks (GNN) exhibit superior performance in graph representation learning, but their inference cost can be high, due to an aggregation operation that can require a…
The Heterophilic Graph Learning Handbook: Benchmarks, Models, Theoretical Analysis, Applications and Challenges
Sitao Luan, Chenqing Hua, Qincheng Lu +11
Homophily principle, \ie{} nodes with the same labels or similar attributes are more likely to be connected, has been commonly believed to be the main reason for the superiority of…
CKGConv: General Graph Convolution with Continuous Kernels
Liheng Ma, Soumyasundar Pal, Yitian Zhang +3
The existing definitions of graph convolution, either from spatial or spectral perspectives, are inflexible and not unified. Defining a general convolution operator in the graph do…
Revealing Decurve Flows for Generalized Graph Propagation
Chen Lin, Liheng Ma, Yiyang Chen +3
This study addresses the limitations of the traditional analysis of message-passing, central to graph learning, by defining {\em \textbf{generalized propagation}} with directed and…
Multi-resolution Time-Series Transformer for Long-term Forecasting
Yitian Zhang, Liheng Ma, Soumyasundar Pal +2
The performance of transformers for time-series forecasting has improved significantly. Recent architectures learn complex temporal patterns by segmenting a time-series into patche…
Graph Inductive Biases in Transformers without Message Passing
Liheng Ma, Chen Lin, Derek Lim +5
Transformers for graph data are increasingly widely studied and successful in numerous learning tasks. Graph inductive biases are crucial for Graph Transformers, and previous works…
Graph Attention Networks with Positional Embeddings
Liheng Ma, Reihaneh Rabbany, Adriana Romero-Soriano
Graph Neural Networks (GNNs) are deep learning methods which provide the current state of the art performance in node classification tasks. GNNs often assume homophily -- neighbori…