Wenwen Qiang

GUPO: Gradient Uncertainty-aware Policy Optimization for Post-Training Large Language Models

Peizheng Guo, Jianqi Zhang, Xingyu Zhang +4

Group Relative Policy Optimization (GRPO) has become a widely used approach for post-training Large Language Models (LLMs) for reasoning. In GRPO, the group gradients induced by di…

Ground-Truth Neighborhood Regularization for Reinforcement Learning Post-Training of Time Series Foundation Models

Jianqi Zhang, Xingyu Zhang, Zeen Song +3

Time series forecasting (TSF) plays an important role in a wide range of real-world applications. Recently, time series foundation models (TSFMs), pretrained on large-scale dataset…

Dirichlet-Guided Group Forecasting for Alleviating Over-smoothing in Time Series Forecasting

Xingyu Zhang, Jingyao Wang, Xin Yu +4

Time series forecasting often suffers from over-smoothing, especially when future dynamics are multi-modal. Forecasts may follow the coarse trend of the observed future, but fail t…

CAMD: Coverage-Aware Multimodal Decoding for Efficient Reasoning of Multimodal Large Language Models

Huijie Guo, Jingyao Wang, Lingyu Si +3

Recent advances in Multimodal Large Language Models (MLLMs) have shown impressive reasoning capabilities across vision-language tasks, yet still face the challenge of compute-diffi…

From Shallow to Deep: Pinning Semantic Intent via Causal GRPO

Shuyi Zhou, Zeen Song, Wenwen Qiang +4

Large Language Models remain vulnerable to adversarial prefix attacks (e.g., ``Sure, here is'') despite robust standard safety. We diagnose this vulnerability as Shallow Safety Ali…

Adaptive Uncertainty-Aware Tree Search for Robust Reasoning

Zeen Song, Zihao Ma, Wenwen Qiang +2

Inference-time reasoning scaling has significantly advanced the capabilities of Large Language Models (LLMs) in complex problem-solving. A prevalent approach involves external sear…

On the Plasticity and Stability for Post-Training Large Language Models

Wenwen Qiang, Ziyin Gu, Jiahuan Zhou +4

Training stability remains a critical bottleneck for Group Relative Policy Optimization (GRPO), often manifesting as a trade-off between reasoning plasticity and general capability…

Towards Generalizable Reasoning: Group Causal Counterfactual Policy Optimization for LLM Reasoning

Jingyao Wang, Peizheng Guo, Wenwen Qiang +4

Large language models (LLMs) excel at complex tasks with advances in reasoning capabilities. However, existing reward mechanisms remain tightly coupled to final correctness and pay…

Enhancing Large Language Models for Time-Series Forecasting via Vector-Injected In-Context Learning

Jianqi Zhang, Jingyao Wang, Wenwen Qiang +2

The World Wide Web needs reliable predictive capabilities to respond to changes in user behavior and usage patterns. Time series forecasting (TSF) is a key means to achieve this go…

Exploring Transferability of Self-Supervised Learning by Task Conflict Calibration

Huijie Guo, Jingyao Wang, Peizheng Guo +3

In this paper, we explore the transferability of SSL by addressing two central questions: (i) what is the representation transferability of SSL, and (ii) how can we effectively mod…

A Generalized Learning Framework for Self-Supervised Contrastive Learning

Lingyu Si, Jingyao Wang, Wenwen Qiang

Self-supervised contrastive learning (SSCL) has recently demonstrated superiority in multiple downstream tasks. In this paper, we generalize the standard SSCL methods to a Generali…

Group Causal Policy Optimization for Post-Training Large Language Models

Ziyin Gu, Jingyao Wang, Ran Zuo +4

Recent advances in large language models (LLMs) have broadened their applicability across diverse tasks, yet specialized domains still require targeted post training. Among existin…

Causal Reward Adjustment: Mitigating Reward Hacking in External Reasoning via Backdoor Correction

Ruike Song, Zeen Song, Huijie Guo +1

External reasoning systems combine language models with process reward models (PRMs) to select high-quality reasoning paths for complex tasks such as mathematical problem solving.…

Reward Model Generalization for Compute-Aware Test-Time Reasoning

Zeen Song, Wenwen Qiang, Siyu Zhao +2

External test-time reasoning enhances large language models (LLMs) by decoupling generation and selection. At inference time, the model generates multiple reasoning paths, and an a…

On the Out-of-Distribution Generalization of Self-Supervised Learning

Wenwen Qiang, Jingyao Wang, Zeen Song +2

In this paper, we focus on the out-of-distribution (OOD) generalization of self-supervised learning (SSL). By analyzing the mini-batch construction during the SSL training phase, w…

Rethinking Multi-Modal Learning from Gradient Uncertainty

Peizheng Guo, Jingyao Wang, Wenwen Qiang +3

Multi-Modal Learning (MML) integrates information from diverse modalities to improve predictive accuracy. While existing optimization strategies have made significant strides by mi…

Beyond All-to-All: Causal-Aligned Transformer with Dynamic Structure Learning for Multivariate Time Series Forecasting

Xingyu Zhang, Hanyun Du, Zeen Song +3

Most existing multivariate time series forecasting methods adopt an all-to-all paradigm that feeds all variable histories into a unified model to predict their future values withou…

Learning to Think: Information-Theoretic Reinforcement Fine-Tuning for LLMs

Jingyao Wang, Wenwen Qiang, Zeen Song +2

Large language models (LLMs) excel at complex tasks thanks to advances in their reasoning abilities. However, existing methods overlook the trade-off between reasoning effectivenes…

Neuromodulated Meta-Learning

Jingyao Wang, Huijie Guo, Wenwen Qiang +4

Humans excel at adapting perceptions and actions to diverse environments, enabling efficient interaction with the external world. This adaptive capability relies on the biological…

Rethinking Meta-Learning from a Learning Lens

Jingyao Wang, Wenwen Qiang, Changwen Zheng +2

Meta-learning seeks to learn a well-generalized model initialization from training tasks to solve unseen tasks. From the "learning to learn" perspective, the quality of the initial…

Towards the Causal Complete Cause of Multi-Modal Representation Learning

Jingyao Wang, Siyu Zhao, Wenwen Qiang +4

Multi-Modal Learning (MML) aims to learn effective representations across modalities for accurate predictions. Existing methods typically focus on modality consistency and specific…

Not All Frequencies Are Created Equal:Towards a Dynamic Fusion of Frequencies in Time-Series Forecasting

Xingyu Zhang, Siyu Zhao, Zeen Song +4

Long-term time series forecasting is a long-standing challenge in various applications. A central issue in time series forecasting is that methods should expressively capture long-…

Teleporter Theory: A General and Simple Approach for Modeling Cross-World Counterfactual Causality

Jiangmeng Li, Bin Qin, Qirui Ji +4

Leveraging the development of structural causal model (SCM), researchers can establish graphical models for exploring the causal mechanisms behind machine learning techniques. As t…