Zhongwei Wan

ClinicalBench: Can LLMs Beat Traditional ML Models in Clinical Prediction?

Canyu Chen, Jian Zhen Yu, Shan Chen, Che Liu et autres

Large Language Models (LLMs) hold great promise to revolutionize current clinical systems for their superior capacities on medical text processing tasks and medical licensing exams. Meanwhile, traditional ML models such as SVM and XGBoost have still been mainly adopted in clinical prediction …

us, gb (code pays fourni par la source)

MathGen: Revealing the Illusion of Mathematical Competence through Text-to-Image Generation

Ruiyao Liu, Hui Shen, Ping Zhang, Yunta Hsieh et autres

Modern generative models have demonstrated the ability to solve challenging mathematical problems. In many real-world settings, however, mathematical solutions must be expressed visually through diagrams, plots, geometric constructions, and structured symbolic layouts, where correctness depends on precise visual composition. This naturally raises …

MathGen: Revealing the Illusion of Mathematical Competence through Text-to-Image Generation

Ruiyao Liu, Hui Shen, Ping Zhang, Yunta Hsieh et autres

Modern generative models have demonstrated the ability to solve challenging mathematical problems. In many real-world settings, however, mathematical solutions must be expressed visually through diagrams, plots, geometric constructions, and structured symbolic layouts, where correctness depends on precise visual composition. This naturally raises …

MMDeepResearch-Bench: A Benchmark for Multimodal Deep Research Agents

Peizhou Huang, Zixuan Zhong, Zhongwei Wan, Donghao Zhou et autres

Deep Research Agents (DRAs) generate citation-rich reports via multi-step search and synthesis, yet existing benchmarks mainly target text-only settings or short-form multimodal QA, missing end-to-end multimodal evidence use. We introduce MMDeepResearch-Bench (MMDR-Bench), a benchmark of 140 expert-crafted tasks across 21 domains, where …

MMDeepResearch-Bench: A Benchmark for Multimodal Deep Research Agents

Peizhou Huang, Zixuan Zhong, Zhongwei Wan, Donghao Zhou et autres

Deep Research Agents (DRAs) generate citation-rich reports via multi-step search and synthesis, yet existing benchmarks mainly target text-only settings or short-form multimodal QA, missing end-to-end multimodal evidence use. We introduce MMDeepResearch-Bench (MMDR-Bench), a benchmark of 140 expert-crafted tasks across 21 domains, where …

MMFormalizer: Multimodal Autoformalization in the Wild

J. Xiong, Qi Han, Yunta Hsieh, Hui Shen et autres

Autoformalization, which translates natural language mathematics into formal statements to enable machine reasoning, faces fundamental challenges in the wild due to the multimodal nature of the physical world, where physics requires inferring hidden constraints (e.g., mass or energy) from visual elements. To …

MMFormalizer: Multimodal Autoformalization in the Wild

J. Xiong, Qi Han, Yunta Hsieh, Hui Shen et autres

Autoformalization, which translates natural language mathematics into formal statements to enable machine reasoning, faces fundamental challenges in the wild due to the multimodal nature of the physical world, where physics requires inferring hidden constraints (e.g., mass or energy) from visual elements. To …

hk, us, gb (code pays fourni par la source)

MedREK: Retrieval-Based Editing for Medical LLMs with Key-Aware Prompts

Shujun Xia, Haokun Lin, Yichen Wu, Zixuan Li et autres

LLMs hold great promise for healthcare applications, but the rapid evolution of medical knowledge and errors in training data often cause them to generate outdated or inaccurate information, limiting their applicability in high-stakes clinical practice. Model editing has emerged as a potential …

Plan Then Action:High-Level Planning Guidance Reinforcement Learning for LLM Reasoning

Qinjian Zhao, Zhongwei Wan, Dinggen Zhang, Weida Wang et autres

Large language models (LLMs) demonstrate strong reasoning abilities via Chain-of-Thought (CoT), but their token-level generation encourages local decisions and lacks global planning, often leading to redundant or inaccurate reasoning. Existing methods, such as tree-based search and reinforcement learning (RL), attempt to address …

ATTS: Asynchronous Test-Time Scaling via Conformal Prediction

Jing Qi Xiong, Qiujiang Chen, Fanghua Ye, Zhongwei Wan et autres

Large language models (LLMs) benefit from test-time scaling but are often hampered by high inference latency. Speculative decoding is a natural way to accelerate the scaling process; however, scaling along both the parallel and sequential dimensions poses significant challenges, including substantial memory-bound …

Enhancing Test-Time Scaling of Large Language Models with Hierarchical Retrieval-Augmented MCTS

Zhongwei Wan, Dongfei Cui, Xin Wang, Jing Xiong et autres

Test-time scaling has emerged as a promising paradigm in language modeling, leveraging additional computational resources at inference time to enhance model performance. In this work, we introduce R2-LLMs, a novel and versatile hierarchical retrieval-augmented reasoning framework designed to improve test-time scaling in …

SwingArena: Competitive Programming Arena for Long-context GitHub Issue Solving

Wendong Xu, Jing Xiong, Chenyang Zhao, Qiujiang Chen et autres

We present SwingArena, a competitive evaluation framework for Large Language Models (LLMs) that closely mirrors real-world software development workflows. Unlike traditional static benchmarks, SwingArena models the collaborative process of software iteration by pairing LLMs as submitters, who generate patches, and reviewers, who …