Yifei Li

NeurIPS Conference 2025 Conference Paper

ChartSketcher: Reasoning with Multimodal Feedback and Reflection for Chart Understanding

• Muye Huang
• Lingling Zhang
• Jie Ma
• Han Lai
• Fangzhi Xu
• Yifei Li
• Wenjun Wu
• Yaqiang Wu

Charts are high-density visualization carriers for complex data, serving as a crucial medium for information extraction and analysis. Automated chart understanding poses significant challenges to existing multimodal large language models (MLLMs) due to the need for precise and complex visual reasoning. Current step-by-step reasoning models primarily focus on text-based logical reasoning for chart understanding. However, they struggle to refine or correct their reasoning when errors stem from flawed visual understanding, as they lack the ability to leverage multimodal interaction for deeper comprehension. Inspired by human cognitive behavior, we propose ChartSketcher, a multimodal feedback-driven step-by-step reasoning method designed to address these limitations. ChartSketcher is a chart understanding model that employs Sketch-CoT, enabling MLLMs to annotate intermediate reasoning steps directly onto charts using a programmatic sketching library, iteratively feeding these visual annotations back into the reasoning process. This mechanism enables the model to visually ground its reasoning and refine its understanding over multiple steps. We employ a two-stage training strategy: a cold start phase to learn sketch-based reasoning patterns, followed by off-policy reinforcement learning to enhance reflection and generalization. Experiments demonstrate that ChartSketcher achieves promising performance on chart understanding benchmarks and general vision tasks, providing an interactive and interpretable approach to chart comprehension.

TCS Journal 2025 Journal Article

Dynamic path finding for multi-load agent pickup and delivery problem

• Yifei Li
• Ruixi Huang
• Hao Ye
• Hejiao Huang
• Hongwei Du

NeurIPS Conference 2025 Conference Paper

HypoBootstrap: A Bootstrapping Framework for Inductive Reasoning

• Si Chen
• Yifei Li
• Richong Zhang

Inductive reasoning infers general rules from observed evidence, which is one of the most critical intelligence abilities. Previous works have succeeded in formal languages but suffer from onerous and error-prone conversions between a particular formal language and the working language. As large language models (LLMs) have emerged, direct reasoning with various kinds of languages, especially natural languages, without formal language involvement has become feasible. However, existing LLM-based inductive reasoning usually relies on LLM's intrinsic generation ability, which is prone to LLM's hallucination and lacks systematic guidance according to the nature of inductive reasoning. To this end, we propose HypoBootstrap, an integrated framework for inductive reasoning that generates and confirms hypotheses both in a bootstrapping manner. Regarding hypothesis generation, we propose a novel bootstrapping generation strategy, bootstrapping object hypotheses, relational hypotheses, and functional hypotheses successively, which assists LLM in observing the evidence from trivial patterns to non-trivial patterns. Regarding hypothesis confirmation, we utilize Glymour's theory of bootstrap confirmation, a hypothesis confirmation theory from the philosophy of science that can confirm a set of hypotheses. We use its principles to confirm the object hypotheses, relational hypotheses, and functional hypotheses. Empirical studies on four inductive reasoning scenarios of different natures, involving causal induction, concept learning, grammar learning, and abstract reasoning, demonstrate that HypoBootstrap significantly outperforms existing methods.

TCS Journal 2025 Journal Article

K-balanced biclique partition: Kernelization and efficient algorithms

• Yifei Li
• Donghua Yang
• Jianzhong Li

IJCAI Conference 2025 Conference Paper

KGCL: Knowledge-Enhanced Graph Contrastive Learning for Retrosynthesis Prediction Based on Molecular Graph Editing

• Fengqin Yang
• Dekui Zhao
• Haoxuan Qiu
• Yifei Li
• Zhiguo Fu

Retrosynthesis, which predicts the reactants of a given target molecule, is an essential task for drug discovery. Retrosynthesis prediction based on molecular graph editing has garnered widespread attention due to excellent interpretability. Existing methods fail to effectively incorporate the chemical knowledge when learning molecular representations. To address this issue, we propose a Knowledge-enhanced Graph Contrastive Learning model (KGCL), which retrieve functional group embeddings from a chemical knowledge graph and integrate them into the atomic embeddings of the product molecule using an attention mechanism. Furthermore, we introduce a graph contrastive learning strategy that generates augmented samples using graph edits to improve the molecular graph encoder. Our proposed method outperforms the strong baseline method Graph2Edits by 1. 6% and 3. 2% in terms of the top-1 accuracy and top-1 round-trip accuracy on the USPTO-50K dataset, respectively, and also achieves a new state-of-the-art performance among semi-template-based methods on the USPTO-FULL dataset.

ICML Conference 2025 Conference Paper

MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding

• Yuxin Zuo
• Shang Qu
• Yifei Li
• Zhang-Ren Chen
• Xuekai Zhu
• Ermo Hua
• Kaiyan Zhang
• Ning Ding 0002

We introduce MedXpertQA, a highly challenging and comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning. MedXpertQA includes 4, 460 questions spanning 17 specialties and 11 body systems. It includes two subsets, Text for text evaluation and MM for multimodal evaluation. Notably, MM introduces expert-level exam questions with diverse images and rich clinical information, including patient records and examination results, setting it apart from traditional medical multimodal benchmarks with simple QA pairs generated from image captions. MedXpertQA applies rigorous filtering and augmentation to address the insufficient difficulty of existing benchmarks like MedQA, and incorporates specialty board questions to improve clinical relevance and comprehensiveness. We perform data synthesis to mitigate data leakage risk and conduct multiple rounds of expert reviews to ensure accuracy and reliability. We evaluate 18 leading models on MedXpertQA. Moreover, medicine is deeply connected to real-world decision-making, providing a rich and representative setting for assessing reasoning abilities beyond mathematics and code. To this end, we develop a reasoning-oriented subset to facilitate the assessment of o1-like models.

NeurIPS Conference 2025 Conference Paper

Mind2Web 2: Evaluating Agentic Search with Agent-as-a-Judge

• Boyu Gou
• Zanming Huang
• Yuting Ning
• Yu Gu
• Michael Lin
• Weijian Qi
• Andrei Kopanev
• Botao Yu

Agentic search such as Deep Research systems-where agents autonomously browse the web, synthesize information, and return comprehensive citation-backed answers-represents a major shift in how users interact with web-scale information. While promising greater efficiency and cognitive offloading, the growing complexity and open-endedness of agentic search have outpaced existing evaluation benchmarks and methodologies, which largely assume short search horizons and static answers. In this paper, we introduce Mind2Web 2, a benchmark of 130 realistic, high-quality, and long-horizon tasks that require real-time web browsing and extensive information synthesis, constructed with over 1000 hours of human labor. To address the challenge of evaluating time-varying and complex answers, we propose a novel Agent-as-a-Judge framework. Our method constructs task-specific judge agents based on a tree-structured rubric design to automatically assess both answer correctness and source attribution. We conduct a comprehensive evaluation of ten frontier agentic search systems and human performance, along with a detailed error analysis to draw insights for future development. The best-performing system, OpenAI Deep Research, can already achieve 50-70% of human performance while spending half the time, highlighting its great potential. Altogether, Mind2Web 2 provides a rigorous foundation for developing and benchmarking the next generation of agentic search systems.

TCS Journal 2025 Journal Article

Task group allocation for multi-load agent pickup and delivery problem

• Yifei Li
• Hao Ye
• Hejiao Huang

ICLR Conference 2024 Conference Paper

Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMs

• Miao Xiong
• Zhiyuan Hu
• Xinyang Lu
• Yifei Li
• Jie Fu 0001
• Junxian He
• Bryan Hooi

Empowering large language models (LLMs) to accurately express confidence in their answers is essential for reliable and trustworthy decision-making. Previous confidence elicitation methods, which primarily rely on *white-box access* to internal model information or model fine-tuning, have become less suitable for LLMs, especially closed-source commercial APIs. This leads to a growing need to explore the untapped area of *black-box* approaches for LLM uncertainty estimation. To better break down the problem, we define a systematic framework with three components: *prompting* strategies for eliciting verbalized confidence, *sampling* methods for generating multiple responses, and *aggregation* techniques for computing consistency. We then benchmark these methods on two key tasks—confidence calibration and failure prediction—across five types of datasets (e.g., commonsense and arithmetic reasoning) and five widely-used LLMs including GPT-4 and LLaMA 2 Chat. Our analysis uncovers several key insights: 1) LLMs, when verbalizing their confidence, tend to be *overconfident*, potentially imitating human patterns of expressing confidence. 2) As model capability scales up, both calibration and failure prediction performance improve, yet still far from ideal performance. 3) Employing our proposed strategies, such as human-inspired prompts, consistency among multiple responses, and better aggregation strategies can help mitigate this overconfidence from various perspectives. 4) Comparisons with white-box methods indicate that while white-box methods perform better, the gap is narrow, e.g., 0.522 to 0.605 in AUROC. Despite these advancements, none of these techniques consistently outperform others, and all investigated methods struggle in challenging tasks, such as those requiring professional knowledge, indicating significant scope for improvement. We believe this study can serve as a strong baseline and provide insights for eliciting confidence in black-box LLMs. The code is publicly available at https://github.com/MiaoXiong2320/llm-uncertainty.

NeurIPS Conference 2024 Conference Paper

NeuralFluid: Nueral Fluidic System Design and Control with Differentiable Simulation

• Yifei Li
• Yuchen Sun
• Pingchuan Ma
• Eftychios Sifakis
• Tao Du
• Bo Zhu
• Wojciech Matusik

We present NeuralFluid, a novel framework to explore neural control and design of complex fluidic systems with dynamic solid boundaries. Our system features a fast differentiable Navier-Stokes solver with solid-fluid interface handling, a low-dimensional differentiable parametric geometry representation, a control-shape co-design algorithm, and gym-like simulation environments to facilitate various fluidic control design applications. Additionally, we present a benchmark of design, control, and learning tasks on high-fidelity, high-resolution dynamic fluid environments that pose challenges for existing differentiable fluid simulators. These tasks include designing the control of artificial hearts, identifying robotic end-effector shapes, and controlling a fluid gate. By seamlessly incorporating our differentiable fluid simulator into a learning framework, we demonstrate successful design, control, and learning results that surpass gradient-free solutions in these benchmark tasks.