16 papers
LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection
Xinhao Zhong, Yuxia Qiao, Junhao Li +3
Reinforcement-learning (RL) post-training equips multimodal large reasoning models (MLRMs) with exploratory chains of thought (CoT), substantially improving visual reasoning. Howev…
FlowErase-OPD: Multi-Concept Erasure via Anchored On-Policy Distillation in Flow Matching Models
Yi Sun, Yimin Zhou, Xinhao Zhong +3
Recent advances in flow matching models have substantially improved the quality of text-to-image generation, but have also raised increasing safety concerns due to their potential…
FlowErase-RL: Rethinking Concept Erasure as Reward Optimization in Flow Matching Models
Yi Sun, Zhiqi Zhang, Xinhao Zhong +5
Recent advances in flow matching models have significantly improved text-to-image generation quality, but also introduce growing safety risks due to the generation of harmful or un…
C2P-VAR: Continual and Compositional Personalization in Visual Autoregressive Models
Junhao Li, Xinhao Zhong, Yi sun +4
Visual autoregressive (VAR) models have recently emerged as an efficient paradigm for text-to-image generation, yet their personalization capabilities remain largely limited to sta…
Mistletoe: Stealthy Acceleration-Collapse Attacks on Speculative Decoding
Shuoyang Sun, Chang Dai, Hao Fang +6
Speculative decoding has become a widely adopted technique for accelerating large language model (LLM) inference by drafting multiple candidate tokens and verifying them with a tar…
SPREG: Structured Plan Repair with Entropy-Guided Test-Time Intervention for Large Language Model Reasoning
Xuan Wang, Yu Ming, Xinhao Zhong +4
Large Language Models (LLMs) are prone to logical hallucinations and stochastic drifts during long-chain reasoning. While Classifier-Free Guidance (CFG) can improve instruction adh…
Differential Vector Erasure: Unified Training-Free Concept Erasure for Flow Matching Models
Zhiqi Zhang, Xinhao Zhong, Yi Sun +4
Text-to-image diffusion models have demonstrated remarkable capabilities in generating high-quality images, yet their tendency to reproduce undesirable concepts, such as NSFW conte…
ActErase: A Training-Free Paradigm for Precise Concept Erasure via Activation Redirection
Yi Sun, Xinhao Zhong, Hongyan Li +4
Recent advances in text-to-image diffusion models have demonstrated remarkable generation capabilities, yet they raise significant concerns regarding safety, copyright, and ethical…
Closing the Safety Gap: Surgical Concept Erasure in Visual Autoregressive Models
Xinhao Zhong, Yimin Zhou, Zhiqi Zhang +6
The rapid progress of visual autoregressive (VAR) models has brought new opportunities for text-to-image generation, but also heightened safety concerns. Existing concept erasure t…
Rectified Decoupled Dataset Distillation: A Closer Look for Fair and Comprehensive Evaluation
Xinhao Zhong, Shuoyang Sun, Xulin Gu +3
Dataset distillation aims to generate compact synthetic datasets that enable models trained on them to achieve performance comparable to those trained on full real datasets, while…
Temporal Saliency-Guided Distillation: A Scalable Framework for Distilling Video Datasets
Xulin Gu, Xinhao Zhong, Zhixing Wei +5
Dataset distillation (DD) has emerged as a powerful paradigm for dataset compression, enabling the synthesis of compact surrogate datasets that approximate the training utility of…
DD-Ranking: Rethinking the Evaluation of Dataset Distillation
Zekai Li, Xinhao Zhong, Samir Khaki +49
In recent years, dataset distillation has provided a reliable solution for data compression, where models trained on the resulting smaller synthetic datasets achieve performance co…
Collaborative Feature-Logits Contrastive Learning for Open-Set Semi-Supervised Object Detection
Xinhao Zhong, Siyu Jiao, Yao Zhao +1
Current Semi-Supervised Object Detection (SSOD) methods enhance detector performance by leveraging large amounts of unlabeled data, assuming that both labeled and unlabeled data sh…
Towards Consistent and Efficient Dataset Distillation via Diffusion-Driven Selection
Xinhao Zhong, Shuoyang Sun, Zhaoyang Xu +4
Dataset distillation provides an effective approach to reduce memory and computational costs by optimizing a compact dataset that achieves performance comparable to the full origin…
Going Beyond Feature Similarity: Effective Dataset Distillation based on Class-Aware Conditional Mutual Information
Xinhao Zhong, Bin Chen, Hao Fang +3
Dataset distillation (DD) aims to minimize the time and memory consumption needed for training deep neural networks on large datasets, by creating a smaller synthetic dataset that…
Hierarchical Features Matter: A Deep Exploration of Progressive Parameterization Method for Dataset Distillation
Xinhao Zhong, Hao Fang, Bin Chen +4
Dataset distillation is an emerging dataset reduction method, which condenses large-scale datasets while maintaining task accuracy. Current parameterization methods achieve enhance…