42 citations · 59 across the 12 of their papers we have counts for
16 papers
Who Gets Unmasked? Sampler-Induced Bias in Diffusion Language Models
Xujun Che, Farsheed Haque, Depeng Xu +2
Diffusion language models (dLLMs) generate by gradually unmasking a canvas, so their samplers decide not only which token to write but also when to commit it. We show that this sec…
CARA: Causally Guided Retrieval Agent for Question Answering
Weicheng Liu, Xiangcheng Wu, Depeng Xu +2
Retrieval-augmented generation over text-rich knowledge graphs requires identifying evidence that contributes to answering a question, rather than merely matching its textual conte…
Forging LLM Authorship Fingerprints with Targeted Rewriting
Haohan Yuan, Simin Chen, Xi Niu +3
Model-attribution classifiers can often identify which language model produced a text, making model-specific writing patterns a signal of provenance. Accurate attribution on unmodi…
Privacy Amplification Without Independence: How Far Negative Dependence Carries the Guarantees of Poisson Subsampling
Xujun Che, Depeng Xu
Poisson subsampling is the default sampler in differentially private optimization because its independence makes privacy amplification tractable. Practical systems, however, are mo…
Memorization Is Not Extraction: Tight Differential-Privacy Bounds and Audit Blind Spots
Xujun Che, Depeng Xu, Shuhan Yuan
Memorization in large language models is measured through a zoo of definitions whose formal relations are unknown, and differential privacy (DP) is treated as a proxy against all o…
MAVISEG: Manifold Propagation and Visual Prototypes for Zero-Shot Open-Vocabulary Segmentation in Diffusion Transformers
Rajatsubhra Chakraborty, Xujun Che, Ritabrata Chakraborty +2
Text-to-image diffusion transformers learn about objects and scenes by learning to generate them, making them strong candidates for training-free zero-shot open-vocabulary semantic…
StraightDP: Geometry-Aware Differential Privacy for Rectified-Flow Transformers
Xujun Che, Depeng Xu, Xintao Wu
Differentially private (DP) training of text-conditioned generative models suffers a utility cliff at strong privacy. We revisit this problem through the geometry of rectified flow…
Unlocking High-Fidelity Molecular Generation from Mass Spectra via Dual-Stream Line Graph Diffusion
Xujun Che, Xiuxia Du, Depeng Xu
De novo molecular generation from tandem mass spectra is a challenging inverse problem whose core difficulty lies in the circular dependency between atom-level and bond-level reaso…
BiasMap: Leveraging Cross-Attentions to Discover and Mitigate Hidden Social Biases in Text-to-Image Generation
Rajatsubhra Chakraborty, Xujun Che, Depeng Xu +3
Bias discovery is critical for black-box generative models, especiall text-to-image (TTI) models. Existing works predominantly focus on output-level demographic distributions, whic…
BadSAD: Clean-Label Backdoor Attacks against Deep Semi-Supervised Anomaly Detection
He Cheng, Depeng Xu, Shuhan Yuan
Image anomaly detection (IAD) is essential in applications such as industrial inspection, medical imaging, and security. Despite the progress achieved with deep learning models lik…
Generating Textual Adversaries with Minimal Perturbation
Xingyi Zhao, Lu Zhang, Depeng Xu +1
Many word-level adversarial attack approaches for textual data have been proposed in recent studies. However, due to the massive search space consisting of combinations of candidat…
Fine-grained Anomaly Detection in Sequential Data via Counterfactual Explanations
He Cheng, Depeng Xu, Shuhan Yuan +1
Anomaly detection in sequential data has been studied for a long time because of its potential in various applications, such as detecting abnormal system behaviors from log data. A…
Fairness-aware Agnostic Federated Learning
Wei Du, Depeng Xu, Xintao Wu +1
Federated learning is an emerging framework that builds centralized machine learning models with training data distributed across multiple devices. Most of the previous works about…
Removing Disparate Impact of Differentially Private Stochastic Gradient Descent on Model Accuracy
Depeng Xu, Wei Du, Xintao Wu
When we enforce differential privacy in machine learning, the utility-privacy trade-off is different w.r.t. each group. Gradient clipping and random noise addition disproportionate…
Achieving Differential Privacy in Vertically Partitioned Multiparty Learning
Depeng Xu, Shuhan Yuan, Xintao Wu
Preserving differential privacy has been well studied under centralized setting. However, it's very challenging to preserve differential privacy under multiparty setting, especiall…
FairGAN: Fairness-aware Generative Adversarial Networks
Depeng Xu, Shuhan Yuan, Lu Zhang +1
Fairness-aware learning is increasingly important in data mining. Discrimination prevention aims to prevent discrimination in the training data before it is used to conduct predict…