Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
AIGQ: An End-to-End Hybrid Generative Architecture for E-commerce Query RecommendationAlibaba · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)Aug 6UniRank: End-to-End Domain-Specific Reranking of Hybrid Text-Image CandidatesAlibaba · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)Aug 6Self-Evolutionary Reinforced Knowledge Distillation for Multi-Modal Tool-Use AgentsAlibaba · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)Aug 6Interpretability in the Era of Large Language Models: Mechanistic Methodology, Empirical Practices, and ApplicationsAlibaba · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)Aug 6DeGRe: Dense-supervised Generative Reranking for RecommendationAlibaba · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)Aug 6Search2Skill: Skill Distillation Beyond Knowledge Boundaries Via Rubric-Based Reinforcement LearningAlibaba · arXiv (Cornell University)Aug 5Spoken Function Calling: A New Perspective on Spoken Language Understanding for Large Audio Language ModelsAlibaba · arXiv (Cornell University)Aug 5STEP-OPD: Rethinking Output Targets and Internal Dynamics in On-Policy Distillation for Diffusion ModelsAlibaba · arXiv (Cornell University)Aug 5Private Direct Preference Optimization for LLM AlignmentAlibaba · arXiv (Cornell University)Aug 5GeoReward: Mitigating Contextual Variable Overestimation in Vision-Language Models for Cross-Market Preference PredictionAlibaba · arXiv (Cornell University)Aug 5State2State: Environment-Derived Mid-Training for LLM AgentsAlibaba · arXiv (Cornell University)Aug 5When Refusal Looks Safe: The Refusal-Cue Shortcut in Safety Guard ModelsAlibaba · arXiv (Cornell University)Aug 4FinProBench: Evaluating Financial AI Agents with Role-Grounded Rubrics Derived from Professional DeliverablesAlibaba · arXiv (Cornell University)Aug 4FinPerMA: A Theory-Informed, Event-Grounded Personalized-Memory Benchmark for LLM AgentsAlibaba · arXiv (Cornell University)Aug 4Towards Robust Tool Use in Agents via Experience-Driven Adaptive GuidanceAlibaba · arXiv (Cornell University)Aug 4DiffImaginE: Imagine to Verify Entity Types with DiffusionAlibaba · arXiv (Cornell University)Aug 4Agentic Reinforcement Learning with Self-Distilled Reward ShapingAlibaba · arXiv (Cornell University)Aug 4SMOPD: Multi-Reward Reinforcement Learning via Specialize-and-Merge Online Policy DistillationAlibaba · arXiv (Cornell University)Aug 4OmniPack: Unified Token Compression for Efficient Omni-modal Large Language ModelsAlibaba · arXiv (Cornell University)Aug 4What to Edit Next: Visually Aligned Image-Editing Follow-Up Suggestions in Conversational SystemsAlibaba · arXiv (Cornell University)Aug 3Crayotter: Learning Long-Horizon Video Editing Agents via Group-Relative Preference BackpropagationAlibaba · arXiv (Cornell University)Aug 3CAVE: Competence-Aware Visual Boundary Evidence Alignment for Video Temporal GroundingAlibaba · arXiv (Cornell University)Aug 3CoEvo-Mem: Co-Evolving Retrieval Policy and Memory Bank for LLM AgentsAlibaba · arXiv (Cornell University)Aug 3Right Answer, Wrong Method: Shortcut Hacking Misleads the Evaluation of LLM Reasoning on Frontier Science BenchmarksAlibaba · arXiv (Cornell University)Aug 3Beyond Solution-Centric Search: Adaptive Inquiry and Knowledge Revision for Autonomous ML EngineeringAlibaba · arXiv (Cornell University)Aug 3