Author name cluster

Ziming Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

15 papers

2 author rows

AAAI Conference 2026 Conference Paper

DeepPhy: Benchmarking Agentic VLMs on Physical Reasoning

Xinrun Xu
Pi Bu
Ye Wang
Börje F. Karlsson
Ziming Wang
Tengtao Song
Qi Zhu
Jun Song

Although Vision Language Models (VLMs) exhibit strong perceptual abilities and impressive visual reasoning, they struggle with attention to detail and precise action planning in complex, dynamic environments, leading to subpar performance. Real-world tasks typically require complex interactions, advanced spatial reasoning, long-term planning, and continuous strategy refinement, usually necessitating understanding the physics rules of the target scenario. However, evaluating these capabilities in real-world scenarios is often prohibitively expensive. To bridge this gap, we introduce DeepPHY, a novel benchmark framework designed to systematically evaluate VLMs' understanding and reasoning about fundamental physical principles through a series of challenging simulated environments. DeepPHY integrates multiple physical reasoning environments of varying difficulty levels and incorporates fine-grained evaluation metrics. Our evaluation finds that even state-of-the-art VLMs struggle to translate descriptive physical knowledge into precise, predictive control.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Fine-Tuned LLMs Know They Don’t Know: A Parameter-Efficient Approach to Recovering Honesty

Zeyu Shi
Ziming Wang
Tianyu Chen
Shiqi Gao
Haoyi Zhou
Qingyun Sun
Jianxin Li

The honesty of Large Language Models (LLMs) is increasingly important for safe deployment in high-stakes domains. However, this crucial trait is severely undermined by supervised fine-tuning (SFT), a common technique for model specialization. Existing recovery methods rely on data-intensive global parameter adjustments, implicitly assuming that SFT deeply corrupts the models' ability to recognize their knowledge boundaries. However, we observe that fine‑tuned LLMs still preserve this ability; what is damaged is their capacity to faithfully express that awareness. Building on this, we propose Honesty-Critical Neurons Restoration (HCNR) to surgically repair this suppressed capacity. HCNR identifies and restores key expression-governing neurons to their pre-trained state while harmonizing them with task-oriented neurons via Hessian-guided compensation. Experiments on four QA tasks and five LLM families demonstrate that HCNR effectively recovers 33.25% of the compromised honesty while achieving at least 2.23x speedup with over 10x less data compared to baseline methods, offering a practical solution for trustworthy LLM deployment.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Global Compression Commander: Plug-and-Play Inference Acceleration for High-Resolution Large Vision-Language Models

Xuyang Liu
Ziming Wang
Junjie Chen
Yuhang Han
Yingyao Wang
Jiale Yuan
Jun Song
Siteng Huang

Large vision-language models (LVLMs) excel at visual understanding but face efficiency challenges due to quadratic complexity when processing long multimodal contexts. While token compression can reduce computational costs, existing approaches are designed for single-view LVLMs and fail to account for the unique multi-view characteristics of high-resolution LVLMs that use dynamic cropping. Current methods treat all tokens uniformly, yet our analysis shows that global thumbnails can naturally guide the compression of local crops by providing holistic context for evaluating informativeness. In this paper, we first analyze the dynamic cropping strategy, revealing both the complementary relationship between thumbnails and crops and the distinct characteristics across different crops. Based on these insights, we propose ''Global Compression Commander'' (GlobalCom2), a novel plug-and-play token compression framework for high-resolution LVLMs. GlobalCom2 uses the thumbnail as a ''commander'' to adaptively guide the compression of local crops, preserving informative details while removing redundancy. Extensive experiments demonstrate that GlobalCom2 maintains over 90% of model performance while compressing 90% of visual tokens, reducing FLOPs to 9.1% and peak memory usage to 60% of the original.

PDF Details DOI

JAIR Journal 2026 Journal Article

Procedural Fairness in Machine Learning

Ziming Wang
Changwu Huang
Ke Tang
Xin Yao

Fairness in machine learning (ML) has garnered significant attention. However, current research has mainly concentrated on the distributive fairness of ML models, with limited focus on another dimension of fairness, i.e., procedural fairness. In this paper, we first define the procedural fairness of ML models by drawing from the established understanding of procedural fairness in philosophy and psychology fields, and then give formal definitions of individual and group procedural fairness. Based on the proposed definition, we further propose a novel metric to evaluate the group procedural fairness of ML models, called GPFFAE, which utilizes a widely used explainable artificial intelligence technique, namely feature attribution explanation (FAE), to capture the decision process of ML models. We validate the effectiveness of GPFFAE on a synthetic dataset and eight real-world datasets. Our experimental studies have revealed the relationship between procedural and distributive fairness of ML models. After validating the proposed metric for assessing the procedural fairness of ML models, we then propose a method for identifying the features that lead to the procedural unfairness of the model and propose two methods to improve procedural fairness based on the identified unfair features. Our experimental results demonstrate that we can accurately identify the features that lead to procedural unfairness in the ML model, and both of our proposed methods can significantly improve procedural fairness while also improving distributive fairness, with a slight sacrifice on the model performance.

PDF Details DOI

IROS Conference 2025 Conference Paper

DISCOVERSE: Efficient Robot Simulation in Complex High-Fidelity Environments

Yufei Jia
Guangyu Wang
Yuhang Dong
Junzhe Wu
Yupei Zeng
Haonan Lin
Zifan Wang
Haizhou Ge

We present Discoverse, the first unified, modular, open-source 3DGS-based simulation framework for Real2Sim2Real robot learning. It features a holistic Real2Sim pipeline that synthesizes hyper-realistic geometry and appearance of complex real-world scenarios, paving the way for analyzing and bridging the Sim2Real gap. Powered by Gaussian Splatting and MuJoCo, Discoverse enables massively parallel simulation of multiple sensor modalities and accurate physics, with inclusive supports for existing 3D assets, robot models, and ROS plugins, empowering large-scale robot learning and complex robotic benchmarks. Through extensive experiments on imitation learning, Dis coverse demonstrates state-of-the-art zero-shot Sim2Real transfer performance compared to existing simulators. For code and demos: https://air-discoverse.github.io/.

Details

AAAI Conference 2025 Conference Paper

GRSN: Gated Recurrent Spiking Neurons for POMDPs and MARL

Lang Qin
Ziming Wang
Runhao Jiang
Rui Yan
Huajin Tang

Spiking neural networks (SNNs) are widely applied in various fields due to their energy-efficient and fast-inference capabilities. Applying SNNs to reinforcement learning (RL) can significantly reduce the computational resource requirements for agents and improve the algorithm's performance under resource-constrained conditions. However, in current spiking reinforcement learning (SRL) algorithms, the simulation results of multiple time steps can only correspond to a single-step decision in RL. This is quite different from the real temporal dynamics in the brain and also fails to fully exploit the capacity of SNNs to process temporal data. In order to address this temporal mismatch issue and further take advantage of the inherent temporal dynamics of spiking neurons, we propose a novel temporal alignment paradigm (TAP) that leverages the single-step update of spiking neurons to accumulate historical state information in RL and introduces gated units to enhance the memory capacity of spiking neurons. Experimental results show that our method can solve partially observable Markov decision processes (POMDPs) and multi-agent cooperation problems with similar performance as recurrent neural networks (RNNs) but with about 50\% power consumption.

PDF Details DOI

ICML Conference 2025 Conference Paper

Theoretical Performance Guarantees for Partial Domain Adaptation via Partial Optimal Transport

Jayadev Naram
Fredrik Hellström
Ziming Wang
Rebecka Jörnsten
Giuseppe Durisi

In many scenarios of practical interest, labeled data from a target distribution are scarce while labeled data from a related source distribution are abundant. One particular setting of interest arises when the target label space is a subset of the source label space, leading to the framework of partial domain adaptation (PDA). Typical approaches to PDA involve minimizing a domain alignment term and a weighted empirical loss on the source data, with the aim of transferring knowledge between domains. However, a theoretical basis for this procedure is lacking, and in particular, most existing weighting schemes are heuristic. In this work, we derive generalization bounds for the PDA problem based on partial optimal transport. These bounds corroborate the use of the partial Wasserstein distance as a domain alignment term, and lead to theoretically motivated explicit expressions for the empirical source loss weights. Inspired by these bounds, we devise a practical algorithm for PDA, termed WARMPOT. Through extensive numerical experiments, we show that WARMPOT is competitive with recent approaches, and that our proposed weights improve on existing schemes.

Details

TAAS Journal 2024 Journal Article

A Roadmap of Explainable Artificial Intelligence: Explain to Whom, When, What and How?

Ziming Wang
Changwu Huang
Xin Yao

Explainable artificial intelligence (XAI) has gained significant attention, especially in AI-powered autonomous and adaptive systems (AASs). However, a discernible disconnect exists among research efforts across different communities. The machine learning community often overlooks “explaining to whom,” while the human-computer interaction community has examined various stakeholders with diverse explanation needs without addressing which XAI methods meet these requirements. Currently, no clear guidance exists on which XAI methods suit which specific stakeholders and their distinct needs. This hinders the achievement of the goal of XAI: providing human users with understandable interpretations. To bridge this gap, this article presents a comprehensive XAI roadmap. Based on an extensive literature review, the roadmap summarizes different stakeholders, their explanation needs at different stages of the AI system lifecycle, the questions they may pose, and existing XAI methods. Then, by utilizing stakeholders’ inquiries as a conduit, the roadmap connects their needs to prevailing XAI methods, providing a guideline to assist researchers and practitioners to determine more easily which XAI methodologies can meet the specific needs of stakeholders in AASs. Finally, the roadmap discusses the limitations of existing XAI methods and outlines directions for future research.

Details DOI

AAAI Conference 2024 Conference Paper

DVSAI: Diverse View-Shared Anchors Based Incomplete Multi-View Clustering

Shengju Yu
Siwei Wang
Pei Zhang
Miao Wang
Ziming Wang
Zhe Liu
Liming Fang
En Zhu

In numerous real-world applications, it is quite common that sample information is partially available for some views due to machine breakdown or sensor failure, causing the problem of incomplete multi-view clustering (IMVC). While several IMVC approaches using view-shared anchors have successfully achieved pleasing performance improvement, (1) they generally construct anchors with only one dimension, which could deteriorate the multi-view diversity, bringing about serious information loss; (2) the constructed anchors are typically with a single size, which could not sufficiently characterize the distribution of the whole samples, leading to limited clustering performance. For generating view-shared anchors with multi-dimension and multi-size for IMVC, we design a novel framework called Diverse View-Shared Anchors based Incomplete multi-view clustering (DVSAI). Concretely, we associate each partial view with several potential spaces. In each space, we enable anchors to communicate among views and generate the view-shared anchors with space-specific dimension and size. Consequently, spaces with various scales make the generated view-shared anchors enjoy diverse dimensions and sizes. Subsequently, we devise an integration scheme with linear computational and memory expenditures to integrate the outputted multi-scale unified anchor graphs such that running spectral algorithm generates the spectral embedding. Afterwards, we theoretically demonstrate that DVSAI owns linear time and space costs, thus well-suited for tackling large-size datasets. Finally, comprehensive experiments confirm the effectiveness and advantages of DVSAI.

PDF Details DOI

IROS Conference 2024 Conference Paper

Ensuring Safety in LLM-Driven Robotics: A Cross-Layer Sequence Supervision Mechanism

Ziming Wang
Qingchen Liu
Jiahu Qin
Man Li 0002

Integrating Large Language Models (LLMs) into robotics significantly enhances autonomous task planning. However, ensuring that multi-step task plans (action sequence) generated by LLMs comply with pre-defined safety constraints during planning and execution remains a challenge, limiting their adaptability in complex environments. To address this issue, a mechanism that can monitor and adjust the plan generated by the LLM-driven task planner and guide the motion planner to avoid potential risks during action execution is required. Therefore, this paper proposes a cross-layer sequence supervision mechanism. Specifically, we employ linear temporal logic syntax to express safety constraints and convert them into a set of nondeterministic Büchi automatons to build a cross-layer safety supervisor. For the task planning layer, the safety supervisor provides a closed-loop correction mechanism that can identify violations in the task plan in real time and guide LLM-driven planners to correct this plan to ensure compliance. For the motion planning layer, the safety supervisor introduces virtual "obstacle" information into the task plan to form the task plan tuple. Based on this plan tuple, the motion planner can proactively prevent unsafe behaviors during action execution. Extensive experimentation demonstrates significant improvements in safety with this cross-layer supervision mechanism, highlighting its potential to enhance LLM-driven robotic technology. Experiment details can be found in https://youtu.be/BDdSSEP6HJw.

Details

NeurIPS Conference 2024 Conference Paper

SE(3)-bi-equivariant Transformers for Point Cloud Assembly

Ziming Wang
Rebecka Jörnsten

Given a pair of point clouds, the goal of assembly is to recover a rigid transformation that aligns one point cloud to the other. This task is challenging because the point clouds may be non-overlapped, and they may have arbitrary initial positions. To address these difficulties, we propose a method, called $SE(3)$-bi-equivariant transformer (BITR), based on the $SE(3)$-bi-equivariance prior of the task: it guarantees that when the inputs are rigidly perturbed, the output will transform accordingly. Due to its equivariance property, BITR can not only handle non-overlapped PCs, but also guarantee robustness against initial positions. Specifically, BITR first extracts features of the inputs using a novel $SE(3) \times SE(3)$-transformer, and then projects the learned feature to group $SE(3)$ as the output. Moreover, we theoretically show that swap and scale equivariances can be incorporated into BITR, thus it further guarantees stable performance under scaling and swapping the inputs. We experimentally show the effectiveness of BITR in practical tasks.

PDF Details DOI

ICML Conference 2023 Conference Paper

Adaptive Smoothing Gradient Learning for Spiking Neural Networks

Ziming Wang
Runhao Jiang
Shuang Lian
Rui Yan 0005
Huajin Tang

Spiking neural networks (SNNs) with biologically inspired spatio-temporal dynamics demonstrate superior energy efficiency on neuromorphic architectures. Error backpropagation in SNNs is prohibited by the all-or-none nature of spikes. The existing solution circumvents this problem by a relaxation on the gradient calculation using a continuous function with a constant relaxation de- gree, so-called surrogate gradient learning. Nevertheless, such a solution introduces additional smoothing error on spike firing which leads to the gradients being estimated inaccurately. Thus, how to adaptively adjust the relaxation degree and eliminate smoothing error progressively is crucial. Here, we propose a methodology such that training a prototype neural network will evolve into training an SNN gradually by fusing the learnable relaxation degree into the network with random spike noise. In this way, the network learns adaptively the accurate gradients of loss landscape in SNNs. The theoretical analysis further shows optimization on such a noisy network could be evolved into optimization on the embedded SNN with shared weights progressively. Moreover, The experiments on static images, dynamic event streams, speech, and instrumental sounds show the proposed method achieves state-of-the-art performance across all the datasets with remarkable robustness on different relaxation degrees.

Details

EAAI Journal 2023 Journal Article

Feature reconstruction graph convolutional network for skeleton-based action recognition

Junhao Huang
Ziming Wang
Jian Peng
Feihu Huang

Skeleton-based action recognition is an important task in computer vision. Recently, graph convolutional networks (GCNs) have been successfully applied to this task and achieved remarkable results. However, there are still some non-negligible limitations with existing GCN-based methods. First, the artificial predefined skeleton partition lacks the joint modeling for different types of edges. Second, most GCN models use interleaved deployment of spatial-only and temporal-only modules to achieve feature learning, which makes them ineffective in capturing spatiotemporal co-occurrence from action sequences. To tackle the above issues, we propose a novel feature reconstruction graph convolutional network (FR-GCN) for skeleton-based action recognition. The proposed FR-GC combines coarse-grained temporal and spatial features to reconstruct fine-grained spatiotemporal features, realizing simultaneous learning of temporal and spatial representations in a single module and significantly improving the capability of the model for spatiotemporal feature extraction. We also propose a topology partition enhancement module to achieve adaptive complementation among different types of edges. Moreover, an efficient multi-scale dual-domain temporal convolution is used to complete further temporal modeling. Compared with state-of-the-art methods, the proposed FR-GCN achieves competitive results on both NTU RGB+D 60 dataset and NTU RGB+D 120 dataset. Especially under the cross-subject benchmarks of the two datasets, the proposed FR-GCN achieves new state-of-the-art performance.

Details DOI

IJCAI Conference 2023 Conference Paper

Learnable Surrogate Gradient for Direct Training Spiking Neural Networks

Shuang Lian
Jiangrong Shen
Qianhui Liu
Ziming Wang
Rui Yan
Huajin Tang

Spiking neural networks (SNNs) have increasingly drawn massive research attention due to biological interpretability and efficient computation. Recent achievements are devoted to utilizing the surrogate gradient (SG) method to avoid the dilemma of non-differentiability of spiking activity to directly train SNNs by backpropagation. However, the fixed width of the SG leads to gradient vanishing and mismatch problems, thus limiting the performance of directly trained SNNs. In this work, we propose a novel perspective to unlock the width limitation of SG, called the learnable surrogate gradient (LSG) method. The LSG method modulates the width of SG according to the change of the distribution of the membrane potentials, which is identified to be related to the decay factors based on our theoretical analysis. Then we introduce the trainable decay factors to implement the LSG method, which can optimize the width of SG automatically during training to avoid the gradient vanishing and mismatch problems caused by the limited width of SG. We evaluate the proposed LSG method on both image and neuromorphic datasets. Experimental results show that the LSG method can effectively alleviate the blocking of gradient propagation caused by the limited width of SG when training deep SNNs directly. Meanwhile, the LSG method can help SNNs achieve competitive performance on both latency and accuracy.

PDF Details DOI

ICLR Conference 2022 Conference Paper

Partial Wasserstein Adversarial Network for Non-rigid Point Set Registration

Ziming Wang
Nan Xue 0001
Ling Lei
Gui-Song Xia

Given two point sets, the problem of registration is to recover a transformation that matches one set to the other. This task is challenging due to the presence of large number of outliers, the unknown non-rigid deformations and the large sizes of point sets. To obtain strong robustness against outliers, we formulate the registration problem as a partial distribution matching (PDM) problem, where the goal is to partially match the distributions represented by point sets in a metric space. To handle large point sets, we propose a scalable PDM algorithm by utilizing the efficient partial Wasserstein-1 (PW) discrepancy. Specifically, we derive the Kantorovich-Rubinstein duality for the PW discrepancy, and show its gradient can be explicitly computed. Based on these results, we propose a partial Wasserstein adversarial network (PWAN), which is able to approximate the PW discrepancy by a neural network, and minimize it by gradient descent. In addition, it also incorporates an efficient coherence regularizer for non-rigid transformations to avoid unrealistic deformations. We evaluate PWAN on practical point set registration tasks, and show that the proposed PWAN is robust, scalable and performs more favorably than the state-of-the-art methods.

Details