Profiles2020-2026ORCID 0000-0003-3890-1894
Jian Liang
University of Chinese Academy of Sciences
31Papers
2Last author
30Unique co-authors
4First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers31
- Mitigating the Safety–Utility Trade-off in LLM Alignment via Adaptive Safe Context LearningAuthor 3 of 6ICML 2026 · 3 citations
- Stop Tracking Me! Proactive Defense Against Attribute Inference Attack in LLMsAuthor 2 of 3ICLR 2026 · 3 citations
- Taming Momentum: Rethinking Optimizer States Through Low-Rank ApproximationAuthor 2 of 5ICLR 2026 · 3 citations
- What If Consensus Lies? Selective-Complementary Reinforcement Learning at Test TimeAuthor 2 of 6ACL 2026 · 3 citations
- Modeling Long-Tail Relations in the Operating Room via In-Context Multimodal LearningAuthor 4 of 6ICML 2026
- Reranker Helps, but Not Enough: Towards Strong Poisoning Attacks Against Retrieval-Augmented GenerationAuthor 2 of 6ICML 2026
- Understanding and Mitigating Spurious Signal Amplification in Test-Time Reinforcement Learning for Math ReasoningAuthor 3 of 7ACL Findings 2026
- USE: A Unified Self-Ensembling Framework for Test-Time Prompt TuningAuthor 2 of 4ICML 2026
- Exploring Vacant Classes in Label-Skewed Federated LearningAuthor 3 of 5AAAI 2025 · 17 citations
- Cooperative Pseudo Labeling for Unsupervised Federated ClassificationAuthor 4 of 6ICCV 2025 · 1 citation
- Uni-Layout: Integrating Human Feedback in Unified Layout Generation and EvaluationAuthor 11 of 11ACM MM 2025 · 1 citation
- Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models?Author 3 of 3CVPR 2025
- LoRA-Pro: Are Low-Rank Adapters Properly Optimized?Author 2 of 5ICLR 2025
- Protecting Model Adaptation from Trojans in the Unlabeled DataAuthor 2 of 5AAAI 2025
- R-TPT: Improving Adversarial Robustness of Vision-Language Models through Test-Time Prompt TuningAuthor 2 of 4CVPR 2025
- A Hard-to-Beat Baseline for Training-free CLIP-based AdaptationAuthor 2 of 6ICLR 2024 · 51 citations
- Realistic Unsupervised CLIP Fine-tuning with Universal Entropy OptimizationAuthor 1 of 5ICML 2024 · 13 citations
- Towards Eliminating Hard Label Constraints in Gradient Inversion AttacksAuthor 2 of 3ICLR 2024 · 7 citations
- Pseudo-Calibration: Improving Predictive Uncertainty Estimation in Unsupervised Domain AdaptationAuthor 2 of 4ICML 2024 · 4 citations
- Backdoor Defense via Test-Time Detecting and RepairingAuthor 2 of 3CVPR 2024
- Connecting the Dots: Collaborative Fine-tuning for Black-Box Vision-Language ModelsAuthor 2 of 5ICML 2024
- TALL: Thumbnail Layout for Deepfake Video DetectionAuthor 2 of 6ICCV 2023 · 133 citations
- Rumor Detection with Diverse Counterfactual EvidenceAuthor 4 of 5KDD 2023 · 22 citations
- Mixed Samples as Probes for Unsupervised Model Selection in Domain AdaptationAuthor 2 of 6NeurIPS 2023 · 14 citations
- Informative Data Mining for One-shot Cross-Domain Semantic SegmentationAuthor 2 of 6ICCV 2023 · 12 citations
- Mind the Label Shift of Augmentation-based Graph OOD GeneralizationAuthor 2 of 3CVPR 2023
- DINE: Domain Adaptation from Single and Multiple Black-box PredictorsAuthor 1 of 9CVPR 2022 · 80 citations
- Are You Stealing My Model? Sample Correlation for Fingerprinting Deep Neural NetworksAuthor 2 of 3NeurIPS 2022 · 57 citations
- No Fear of Heterogeneity: Classifier Calibration for Federated Learning with Non-IID DataAuthor 5 of 6NeurIPS 2021 · 510 citations
- Domain Adaptation With Auxiliary Target Domain-Oriented ClassifierAuthor 1 of 3CVPR 2021
Showing the 30 most recent indexed papers.
