Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
Toward Robust Low-Latency Live Streaming: Measurement, Prediction, and Rate Adaptation Under UncertaintyAlibaba · IEEE Transactions on Mobile ComputingJan 20SceneVTG++: Controllable Multilingual Visual Text Generation in the WildAlibaba · IEEE Transactions on Circuits and Systems for Video TechnologyJan 20ThinkFL: Self-Refining Failure Localization for Microservice Systems via Reinforcement Fine-TuningAlibaba · ACM Transactions on Software Engineering and Methodology10 citationsJan 19Probabilistic modeling of disparity uncertainty for robust and efficient stereo matchingAlibaba · Pattern RecognitionJan 18DyDiT++: Diffusion Transformers With Timestep and Spatial Dynamics for Efficient Visual GenerationAlibaba · IEEE Transactions on Pattern Analysis and Machine Intelligence3 citationsJan 15Long-Chain Reasoning Distillation via Adaptive Prefix AlignmentAlibaba · arXiv (Cornell University)Jan 15Evidence-Augmented Policy Optimization with Reward Co-Evolution for Long-Context ReasoningAlibaba · arXiv (Cornell University)Jan 15DecisionLLM: Large Language Models for Long Sequence Decision ExplorationAlibaba · arXiv (Cornell University)Jan 15Long-Chain Reasoning Distillation via Adaptive Prefix AlignmentAlibaba · arXiv (Cornell University)Jan 15Evidence-Augmented Policy Optimization with Reward Co-Evolution for Long-Context ReasoningAlibaba · arXiv (Cornell University)Jan 15LatencyPrism: Online Non-intrusive Latency Sculpting for SLO-Guaranteed LLM InferenceAlibaba · arXiv (Cornell University)Jan 14Distribution-Aligned Sequence Distillation for Superior Long-CoT ReasoningAlibaba · arXiv (Cornell University)Jan 14Triplets Better Than Pairs: Towards Stable and Effective Self-Play Fine-Tuning for LLMsAlibaba · arXiv (Cornell University)Jan 13Triplets Better Than Pairs: Towards Stable and Effective Self-Play Fine-Tuning for LLMsAlibaba · arXiv (Cornell University)Jan 13AugMMRev: An LLM-Augmented Multimodal Ranking Model for Personalized Image Material RetrievalAlibaba · IEEE Transactions on Consumer Electronics8 citationsJan 12SAD: A Large-Scale Strategic Argumentative Dialogue DatasetAlibaba · arXiv (Cornell University)Jan 12Smooth Operator: Smooth Verifiable Reward Activates Spatial Reasoning Ability of Vision-Language ModelAlibaba · arXiv (Cornell University)Jan 12Reward Modeling from Natural Language Human FeedbackAlibaba · arXiv (Cornell University)Jan 12Rewarding Creativity: A Human-Aligned Generative Reward Model for Reinforcement Learning in StorytellingAlibaba · arXiv (Cornell University)Jan 12SAD: A Large-Scale Strategic Argumentative Dialogue DatasetAlibaba · arXiv (Cornell University)Jan 12MegaFlow: Large-Scale Distributed Orchestration System for the Agentic EraAlibaba · arXiv (Cornell University)Jan 12No More Stale Feedback: Co-Evolving Critics for Open-World Agent LearningAlibaba · arXiv (Cornell University)Jan 11No More Stale Feedback: Co-Evolving Critics for Open-World Agent LearningAlibaba · arXiv (Cornell University)Jan 11MedGround: Bridging the Evidence Gap in Medical Vision-Language Models with Verified Grounding DataAlibaba · arXiv (Cornell University)Jan 11ArenaRL: Scaling RL for Open-Ended Agents via Tournament-based Relative RankingAlibaba · arXiv (Cornell University)Jan 10