Profiles2024-2026
Jun Yu
Harbin Institute of Technology
26Papers
14Last author
60Unique co-authors
0First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers26
- Prune Redundancy, Preserve Essence: Vision Token Compression in VLMs via Synergistic Importance-DiversityAuthor 5 of 6ICLR 2026 · 25 citations
- Uni-X: Mitigating Modality Conflict with a Two-End-Separated Architecture for Unified Multimodal ModelsAuthor 5 of 5ICLR 2026 · 18 citations
- Cut to the Chase: Training-free Multimodal Summarization via Chain-of-EventsAuthor 5 of 5CVPR 2026 · 4 citations
- Partially Shared Concept Bottleneck ModelsAuthor 5 of 5AAAI 2026 · 2 citations
- Knowledge Completes the Vision: A Multimodal Entity-aware Retrieval-Augmented Generation Framework for News Image CaptioningAuthor 7 of 7AAAI 2026 · 1 citation
- Measuring Social Bias in Vision-Language Models with Face-Only Counterfactuals from Real PhotosAuthor 6 of 6ACL 2026 · 1 citation
- DiffTrans: Differentiable Geometry-Materials Decomposition for Reconstructing Transparent ObjectsAuthor 5 of 6ICLR 2026
- Memory-Augmented Scene Understanding and Exploration for Open-World Aerial Object-Goal NavigationAuthor 6 of 7CVPR 2026
- PromptDepth: Efficient and Promptable Geometric 3D Vision Model for Embodied IntelligenceAuthor 9 of 9CVPR 2026
- A Token is Worth over 1, 000 Tokens: Efficient Knowledge Distillation through Low-Rank CloneAuthor 6 of 6NeurIPS 2025 · 17 citations
- Benchmarking and Improving Large Vision-Language Models for Fundamental Visual Graph Understanding and ReasoningAuthor 5 of 6ACL 2025 · 15 citations
- EditInfinity: Image Editing with Binary-Quantized Generative ModelsAuthor 3 of 5NeurIPS 2025 · 6 citations
- Fine-grained Adaptive Visual Prompt for Generative Medical Visual Question AnsweringAuthor 3 of 4AAAI 2025 · 6 citations
- A Similarity Paradigm Through Textual Regularization Without ForgettingAuthor 5 of 5AAAI 2025 · 4 citations
- Growing a Twig to Accelerate Large Vision-Language ModelsAuthor 9 of 9ICCV 2025 · 3 citations
- PRISM: A Framework for Producing Interpretable Political Bias Embeddings with Political-Aware Cross-EncoderAuthor 4 of 4ACL 2025 · 2 citations
- Enhancing LLM Planning for Robotics Manipulation through Hierarchical Procedural Knowledge GraphsAuthor 4 of 4NeurIPS 2025 · 1 citation
- A General Framework for Producing Interpretable Semantic Text EmbeddingsAuthor 5 of 5ICLR 2025
- APT: Improving Specialist LLM Performance with Weakness Case Acquisition and Iterative Preference TrainingAuthor 8 of 9ACL Findings 2025
- AQuilt: Weaving Logic and Self-Inspection into Low-Cost, High-Relevance Data Synthesis for Specialist LLMsAuthor 6 of 7EMNLP 2025
- Deep Kernel Relative Test for Machine-generated Text DetectionAuthor 5 of 6ICLR 2025
- Enhancing Target-unspecific Tasks through a Features MatrixAuthor 5 of 5ICML 2025
- MTSA: Multi-turn Safety Alignment for LLMs through Multi-round Red-teamingAuthor 6 of 7ACL 2025
- Recognition-Synergistic Scene Text EditingAuthor 5 of 7CVPR 2025
- SeaPO: Strategic Error Amplification for Robust Preference Optimization of Large Language ModelsAuthor 8 of 9EMNLP Findings 2025
- MVPbev: Multi-view Perspective Image Generation from BEV with Test-time Controllability and GeneralizabilityAuthor 6 of 6ACM MM 2024 · 5 citations
