Author name cluster

Zekai 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.

9 papers

2 author rows

ICRA Conference 2025 Conference Paper

In-Context Learning Enables Robot Action Prediction in LLMs

Yida Yin
Zekai Wang
Yuvan Sharma
Dantong Niu
Trevor Darrell
Roei Herzig

Recently, Large Language Models (LLMs) have achieved remarkable success using in-context learning (ICL) in the language domain. However, leveraging the ICL capabilities within LLMs to directly predict robot actions remains largely unexplored. In this paper, we introduce RoboPrompt, a frame-work that enables off-the-shelf text-only LLMs to directly predict robot actions through ICL without training. Our approach first heuristically identifies keyframes that capture important moments from an episode. Next, we extract end-effector actions from these keyframes as well as the estimated initial object poses, and both are converted into textual descriptions. Finally, we construct a structured template to form ICL demonstrations from these textual descriptions and a task instruction. This enables an LLM to directly predict robot actions at test time. Through extensive experiments and analysis, RoboPrompt shows stronger performance over zero-shot and ICL baselines in simulated and real-world settings. Our project page is available at https://davidyyd.github.io/roboprompt.

Details

AAAI Conference 2024 Conference Paper

DRF: Improving Certified Robustness via Distributional Robustness Framework

Zekai Wang
Zhengyu Zhou
Weiwei Liu

Randomized smoothing (RS) has provided state-of-the-art (SOTA) certified robustness against adversarial perturbations for large neural networks. Among studies in this field, methods based on adversarial training (AT) achieve remarkably robust performance by applying adversarial examples to construct the smoothed classifier. These AT-based RS methods typically seek a pointwise adversary that generates the worst-case adversarial examples by perturbing each input independently. However, there are unexplored benefits to considering such adversarial robustness across the entire data distribution. To this end, we provide a novel framework called DRF, which connects AT-based RS methods with distributional robustness (DR), and show that these methods are special cases of their counterparts in our framework. Due to the advantages conferred by DR, our framework can control the trade-off between the clean accuracy and certified robustness of smoothed classifiers to a significant extent. Our experiments demonstrate that DRF can substantially improve the certified robustness of AT-based RS.

PDF Details DOI

JBHI Journal 2024 Journal Article

Multi-Branching Temporal Convolutional Network With Tensor Data Completion for Diabetic Retinopathy Prediction

Zekai Wang
Suhao Chen
Tieming Liu
Bing Yao

Diabetic retinopathy (DR), a microvascular complication of diabetes, is the leading cause of vision loss among working-aged adults. However, due to the low compliance rate of DR screening and expensive medical devices for ophthalmic exams, many DR patients did not seek proper medical attention until DR develops to irreversible stages (i. e. , vision loss). Fortunately, the widely available electronic health record (EHR) databases provide an unprecedented opportunity to develop cost-effective machine-learning tools for DR detection. This paper proposes a Multi-branching Temporal Convolutional Network with Tensor Data Completion (MB-TCN-TC) model to analyze the longitudinal EHRs collected from diabetic patients for DR prediction. Experimental results demonstrate that the proposed MB-TCN-TC model not only effectively copes with the imbalanced data and missing value issues commonly seen in EHR datasets but also captures the temporal correlation and complicated interactions among medical variables in the longitudinal clinical records, yielding superior prediction performance compared to existing methods. Specifically, our MB-TCN-TC model provides AUROC and AUPRC scores of 0. 949 and 0. 793 respectively, achieving an improvement of 6. 27% on AUROC, 11. 85% on AUPRC, and 19. 3% on F1 score compared with the traditional TCN model.

Details DOI

ICML Conference 2023 Conference Paper

Better Diffusion Models Further Improve Adversarial Training

Zekai Wang
Tianyu Pang
Chao Du
Min Lin
Weiwei Liu 0003
Shuicheng Yan

It has been recognized that the data generated by the denoising diffusion probabilistic model (DDPM) improves adversarial training. After two years of rapid development in diffusion models, a question naturally arises: can better diffusion models further improve adversarial training? This paper gives an affirmative answer by employing the most recent diffusion model which has higher efficiency ($\sim 20$ sampling steps) and image quality (lower FID score) compared with DDPM. Our adversarially trained models achieve state-of-the-art performance on RobustBench using only generated data (no external datasets). Under the $\ell_\infty$-norm threat model with $\epsilon=8/255$, our models achieve $70. 69\\%$ and $42. 67\\%$ robust accuracy on CIFAR-10 and CIFAR-100, respectively, i. e. improving upon previous state-of-the-art models by $+4. 58\\%$ and $+8. 03\\%$. Under the $\ell_2$-norm threat model with $\epsilon=128/255$, our models achieve $84. 86\\%$ on CIFAR-10 ($+4. 44\\%$). These results also beat previous works that use external data. We also provide compelling results on the SVHN and TinyImageNet datasets. Our code is at https: //github. com/wzekai99/DM-Improves-AT.

Details

JMLR Journal 2023 Journal Article

RVCL: Evaluating the Robustness of Contrastive Learning via Verification

Zekai Wang
Weiwei Liu

Contrastive adversarial training has successfully improved the robustness of contrastive learning (CL). However, the robustness metric in these methods depends on attack algorithms, image labels, and downstream tasks, introducing reliability concerns. To address these issues, this paper proposes a novel Robustness Verification framework for Contrastive Learning (RVCL). Specifically, we define the verification problem of CL from deterministic and probabilistic perspectives, then provide several effective metrics to evaluate the robustness of CL encoder. Furthermore, we use extreme value theory to reveal the relationship between the robust radius of the CL encoder and that of the supervised downstream task. Extensive experiments on various benchmark models and datasets validate theoretical findings, and further demonstrate RVCL's capability to evaluate the robustness of both CL encoders and images. Our code is available at https://github.com/wzekai99/RVCL-JMLR. [abs] [ pdf ][ bib ] [ code ] &copy JMLR 2023. ( edit, beta )

PDF Details

JBHI Journal 2022 Journal Article

Multi-Branching Temporal Convolutional Network for Sepsis Prediction

Zekai Wang
Bing Yao

Sepsisis among the leading causes of morbidity and mortality in modern intensive care units. Accurate sepsis prediction is of critical importance to save lives and reduce medical costs. The rapid advancements in sensing and information technology facilitate the effective monitoring of patients’ health conditions, generating a wealth of medical data, and provide an unprecedented opportunity for data-driven diagnosis of sepsis. However, real-world medical data are often complexly structured with a high level of uncertainty (e. g. , missing values, imbalanced data). Realizing the full data potential depends on developing effective analytical models. In this paper, we propose a novel predictive framework with Multi-Branching Temporal Convolutional Network (MB-TCN) to model the complexly structured medical data for robust prediction of sepsis. The MB-TCN framework not only efficiently handles the missing value and imbalanced data issues but also effectively captures the temporal pattern and heterogeneous variable interactions. We evaluate the performance of the proposed MB-TCN in predicting sepsis using real-world medical data from PhysioNet/Computing in Cardiology Challenge 2019. Experimental results show that MB-TCN outperforms existing methods that are commonly used in current practice.

Details DOI

NeurIPS Conference 2022 Conference Paper

On the Tradeoff Between Robustness and Fairness

Xinsong Ma
Zekai Wang
Weiwei Liu

Interestingly, recent experimental results [2, 26, 22] have identified a robust fairness phenomenon in adversarial training (AT), namely that a robust model well-trained by AT exhibits a remarkable disparity of standard accuracy and robust accuracy among different classes compared with natural training. However, the effect of different perturbation radii in AT on robust fairness has not been studied, and one natural question is raised: does a tradeoff exist between average robustness and robust fairness? Our extensive experimental results provide an affirmative answer to this question: with an increasing perturbation radius, stronger AT will lead to a larger class-wise disparity of robust accuracy. Theoretically, we analyze the class-wise performance of adversarially trained linear models with mixture Gaussian distribution. Our theoretical results support our observations. Moreover, our theory shows that adversarial training easily leads to more serious robust fairness issue than natural training. Motivated by theoretical results, we propose a fairly adversarial training (FAT) method to mitigate the tradeoff between average robustness and robust fairness. Experimental results validate the effectiveness of our proposed method.

PDF Details

ICML Conference 2022 Conference Paper

Robustness Verification for Contrastive Learning

Zekai Wang
Weiwei Liu 0003

Contrastive adversarial training has successfully improved the robustness of contrastive learning (CL). However, the robustness metric used in these methods is linked to attack algorithms, image labels and downstream tasks, all of which may affect the consistency and reliability of robustness metric for CL. To address these problems, this paper proposes a novel Robustness Verification framework for Contrastive Learning (RVCL). Furthermore, we use extreme value theory to reveal the relationship between the robust radius of the CL encoder and that of the supervised downstream task. Extensive experimental results on various benchmark models and datasets verify our theoretical findings, and further demonstrate that our proposed RVCL is able to evaluate the robustness of both models and images. Our code is available at https: //github. com/wzekai99/RVCL.

Details

IJCAI Conference 2019 Conference Paper

Unified Embedding Model over Heterogeneous Information Network for Personalized Recommendation

Zekai Wang
Hongzhi Liu
Yingpeng Du
Zhonghai Wu
Xing Zhang

Most of heterogeneous information network (HIN) based recommendation models are based on the user and item modeling with meta-paths. However, they always model users and items in isolation under each meta-path, which may lead to information extraction misled. In addition, they only consider structural features of HINs when modeling users and items during exploring HINs, which may lead to useful information for recommendation lost irreversibly. To address these problems, we propose a HIN based unified embedding model for recommendation, called HueRec. We assume there exist some common characteristics under different meta-paths for each user or item, and use data from all meta-paths to learn unified users’ and items’ representations. So the interrelation between meta-paths are utilized to alleviate the problems of data sparsity and noises on one meta-path. Different from existing models which first explore HINs then make recommendations, we combine these two parts into an end-to-end model to avoid useful information lost in initial phases. In addition, we embed all users, items and meta-paths into related latent spaces. Therefore, we can measure users’ preferences on meta-paths to improve the performances of personalized recommendation. Extensive experiments show HueRec consistently outperforms state-of-the-art methods.

PDF Details