Profiles2020-2026ORCID 0000-0002-7438-5180
Wei Chen
Microsoft Research Asia
48Papers
19Last author
74Unique co-authors
4First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers48
- Benefits and Pitfalls of Reinforcement Learning for Language Model Planning: A Theoretical PerspectiveAuthor 8 of 8ICLR 2026 · 7 citations
- Toward Consistent World Models with Multi-Token Prediction and Latent Semantic EnhancementAuthor 6 of 7ACL 2026 · 3 citations
- Embedding-Space Orthogonal Decomposition for Robust Social RecommendationAuthor 3 of 8KDD 2026 · 1 citation
- RaPA: Enhancing Transferable Targeted Attacks via Random Parameter PruningAuthor 4 of 5CVPR 2026 · 1 citation
- Extra-Merge: Tracing the Rank-1 Subspace of Model Merging in Language Model Pre-TrainingAuthor 5 of 6ICML 2026
- MemFlow: A Lightweight Forward Memorizing Framework for Quick Domain Adaptive Feature MappingAuthor 5 of 6CVPR 2026
- PaT: Planning-after-Trial for Efficient Test-Time Code GenerationAuthor 5 of 6ACL 2026
- The Stability of Singular Distribution: A Spectral Perspective on the Two-Phase Dynamics of Language Model Pre-trainingAuthor 4 of 5ICML 2026
- Mechanism Design for LLM Fine-tuning with Multiple Reward ModelsAuthor 5 of 6NeurIPS 2025 · 26 citations
- CLIPure: Purification in Latent Space via CLIP for Adversarially Robust Zero-Shot ClassificationAuthor 3 of 5ICLR 2025
- NAPPure: Adversarial Purification for Robust Image Classification Under Non-Additive PerturbationsAuthor 3 of 5ICCV 2025
- Offline Learning for Combinatorial Multi-armed BanditsAuthor 7 of 7ICML 2025
- ALPINE: Unveiling The Planning Capability of Autoregressive Learning in Language ModelsAuthor 6 of 6NeurIPS 2024 · 18 citations
- Generative Retrieval Meets Multi-Graded RelevanceAuthor 5 of 6NeurIPS 2024 · 18 citations
- Perturbation-Invariant Adversarial Training for Neural Ranking Models: Improving the Effectiveness-Robustness Trade-OffAuthor 4 of 7AAAI 2024 · 17 citations
- Query-Efficient Correlation Clustering with Noisy OracleAuthor 4 of 4NeurIPS 2024 · 11 citations
- Combinatorial Multivariant Multi-Armed Bandits with Applications to Episodic Reinforcement Learning and BeyondAuthor 10 of 10ICML 2024 · 9 citations
- Provable Adaptivity of Adam under Non-uniform SmoothnessAuthor 5 of 5KDD 2024 · 4 citations
- Cascading Reinforcement LearningAuthor 3 of 3ICLR 2024 · 2 citations
- Combinatorial Bandits for Maximum Value Reward Function under Value-Index FeedbackAuthor 2 of 3ICLR 2024
- Contextual Combinatorial Bandits with Probabilistically Triggered ArmsAuthor 7 of 7ICML 2023 · 26 citations
- Topic-oriented Adversarial Attacks against Black-box Neural Ranking ModelsAuthor 5 of 7SIGIR 2023 · 20 citations
- Scalable Fair Influence MaximizationAuthor 5 of 5NeurIPS 2023 · 17 citations
- Combinatorial Causal BanditsAuthor 2 of 2AAAI 2023 · 16 citations
- Multi-Fidelity Multi-Armed Bandits RevisitedAuthor 3 of 4NeurIPS 2023 · 8 citations
- Popularity Ratio Maximization: Surpassing Competitors through Influence PropagationAuthor 7 of 7SIGMOD 2023 · 6 citations
- Mitigating Filter Bubbles Under a Competitive Diffusion ModelAuthor 2 of 4SIGMOD 2023 · 4 citations
- Closing the gap between the upper bound and lower bound of Adam's iteration complexityAuthor 5 of 5NeurIPS 2023 · 2 citations
- Collaborative Pure Exploration in Kernel BanditAuthor 2 of 4ICLR 2023
- Combinatorial Pure Exploration of Causal BanditsAuthor 2 of 2ICLR 2023
Showing the 30 most recent indexed papers.
