Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
Chart2Code53: A Large-Scale Diverse and Complex Dataset for Enhancing Chart-to-Code GenerationiFlytek · Conference on Empirical Methods in Natural Language Processing (EMNLP)1 citationJan 1, 2025AdaptFlow: Adaptive Workflow Optimization via Meta-LearningMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025Mixture of Languages: Improved Multilingual Encoders Through Language GroupingMeta · Conference on Empirical Methods in Natural Language Processing (EMNLP)1 citationJan 1, 2025NL2Lean: Translating Natural Language into Lean 4 through Multi-Aspect Reinforcement LearningMicrosoft · Conference on Empirical Methods in Natural Language Processing (EMNLP)1 citationJan 1, 2025Reinforcement Learning with Supervised AlignmentMeta · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025iKnow-audio: Integrating Knowledge Graphs with Audio-Language ModelsNVIDIA · Conference on Empirical Methods in Natural Language Processing (EMNLP)1 citationJan 1, 2025ReAlign: Structured Revision for Small Language Model AlignmentMeta · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025Teaching LLMs to Plan, Not Just Solve: Plan Learning Boosts LLMs Generalization in Reasoning TasksMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025SalaMAnder: Shapley-based Mathematical Expression Attribution and Metric for Chain-of-Thought ReasoningAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025CultureSynth: A Hierarchical Taxonomy-Guided and Retrieval-Augmented Framework for Cultural Question-Answer SynthesisAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025ViFT: Towards Visual Instruction-Free Fine-tuning for Large Vision-Language ModelsAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025LongWeave: A Long-Form Generation Benchmark Bridging Real-World Relevance and VerifiabilityAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025Investigating Dictionary Expansion for Video-based Sign Language DictionariesMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025Hierarchical Reward Modeling for Fault Localization in Large Code RepositoriesMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025Intelligent Document Parsing: Towards End-to-end Document Parsing via Decoupled Content Parsing and Layout GroundingAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025Nexus: Adaptive Upcycling to Efficiently Pretrain Mixture of ExpertsCohere · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)1 citationJan 1, 2025Evaluating WMT 2025 Metrics Shared Task Submissions on the SSA-MTE African Challenge SetAlphabet · Conference on Machine Translation (WMT)1 citationJan 1, 2025What’s in Common? Multimodal Models Hallucinate When Reasoning Across ScenesMeta · neural information processing systems1 citationJan 1, 2025Zero-Shot Performance Prediction for Probabilistic Scaling LawsAlphabet · neural information processing systems1 citationJan 1, 2025Infinite Neural Operators: Gaussian processes on functionsGoogle DeepMind · neural information processing systems1 citationJan 1, 2025Heavy-Ball Momentum Method in Continuous Time and Discretization Error AnalysisAlibaba · neural information processing systems1 citationJan 1, 2025Reverse Engineering Human Preferences with Reinforcement LearningCohere, Google DeepMind · neural information processing systems1 citationJan 1, 2025Knowledge-Decoupled Functionally Invariant Path With Synthetic Personal Data for Personalized ASRAlibaba · IEEE Signal Processing Letters1 citationJan 1, 2025Agnostic Learning under Targeted Poisoning: Optimal Rates and the Role of RandomnessAlphabet · neural information processing systems1 citationJan 1, 2025WildCAT3D: Appearance-Aware Multi-View Diffusion in the WildMeta · neural information processing systems1 citationJan 1, 2025