Picture for Gaowei Wu

Gaowei Wu

StructVRM: Aligning Multimodal Reasoning with Structured and Verifiable Reward Models

Add code
Aug 07, 2025
Viaarxiv icon

ChartMind: A Comprehensive Benchmark for Complex Real-world Multimodal Chart Question Answering

Add code
May 29, 2025
Viaarxiv icon

Adapting Step-size: A Unified Perspective to Analyze and Improve Gradient-based Methods for Adversarial Attacks

Add code
Feb 02, 2023
Figure 1 for Adapting Step-size: A Unified Perspective to Analyze and Improve Gradient-based Methods for Adversarial Attacks
Figure 2 for Adapting Step-size: A Unified Perspective to Analyze and Improve Gradient-based Methods for Adversarial Attacks
Figure 3 for Adapting Step-size: A Unified Perspective to Analyze and Improve Gradient-based Methods for Adversarial Attacks
Figure 4 for Adapting Step-size: A Unified Perspective to Analyze and Improve Gradient-based Methods for Adversarial Attacks
Viaarxiv icon

The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods

Add code
Feb 15, 2021
Figure 1 for The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods
Figure 2 for The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods
Figure 3 for The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods
Figure 4 for The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods
Viaarxiv icon