Profiles2024-2026
Wei Liu
MiLM Plus, Xiaomi Inc.
16Papers
2Last author
36Unique co-authors
0First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers16
- End-to-End Optimization of LLM-Driven Multi-Agent Search Systems via Heterogeneous-Group-Based Reinforcement LearningAuthor 4 of 6ACL 2026 · 8 citations
- VecInfer: Efficient LLM Inference with Low-Bit KV Cache via Outlier-Suppressed Vector QuantizationAuthor 5 of 7ACL 2026 · 4 citations
- FutureMind: Equipping Small Language Models with Strategic Thinking-Pattern Priors via Adaptive Knowledge DistillationAuthor 5 of 6ICLR 2026
- MobileBench-OL: A Comprehensive Chinese Benchmark for Evaluating Mobile GUI Agents in Real-World EnvironmentAuthor 5 of 6ACL Findings 2026
- STEP: Success-Rate-Aware Trajectory-Efficient Policy OptimizationAuthor 6 of 6ACL Findings 2026
- More is not always better? Enhancing Many-Shot In-Context Learning with Differentiated and Reweighting ObjectivesAuthor 5 of 11ACL 2025 · 9 citations
- Browsing Like Human: A Multimodal Web Agent with Experiential Fast-and-Slow ThinkingAuthor 3 of 6ACL 2025 · 7 citations
- MobileSteward: Integrating Multiple App-Oriented Agents with Self-Evolution to Automate Cross-App InstructionsAuthor 2 of 5KDD 2025 · 4 citations
- TailorKV: A Hybrid Framework for Long-Context Inference via Tailored KV Cache OptimizationAuthor 4 of 8ACL Findings 2025 · 2 citations
- BacktrackAgent: Enhancing GUI Agent with Error Detection and Backtracking MechanismAuthor 3 of 4EMNLP 2025 · 1 citation
- HoPE: A Novel Positional Encoding Without Long-Term Decay for Enhanced Context Awareness and ExtrapolationAuthor 5 of 5ACL 2025
- SPO: Self Preference Optimization with Self RegularizationAuthor 5 of 6EMNLP Findings 2025
- A Comprehensive Evaluation of Quantization Strategies for Large Language ModelsAuthor 4 of 7ACL Findings 2024 · 55 citations
- MobileVLM: A Vision-Language Model for Better Intra- and Inter-UI UnderstandingAuthor 3 of 9EMNLP Findings 2024 · 22 citations
- Pruning Large Language Models to Intra-module Low-rank Architecture with Transitional ActivationsAuthor 4 of 7ACL Findings 2024 · 3 citations
- ToolPlanner: A Tool Augmented LLM for Multi Granularity Instructions with Path Planning and FeedbackAuthor 2 of 4EMNLP 2024 · 3 citations
