Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
Fine-grained Controllable Video Generation via Object Appearance and ContextGoogle DeepMind · IEEE Winter Conference on Applications of Computer Vision2 citationsFeb 26, 2025FSPO: Few-Shot Optimization of Synthetic Preferences Personalizes to Real UsersAlphabet, Google DeepMind, OpenAI · arXiv (Cornell University)Feb 26, 2025Generating Long-Take Videos via Effective Keyframes and GuidanceGoogle DeepMind · IEEE Winter Conference on Applications of Computer VisionFeb 26, 2025HCI for AGIGoogle DeepMind · interactions10 citationsFeb 25, 2025Optimization of Rank Losses for Image RetrievalGoogle DeepMind · IEEE Transactions on Pattern Analysis and Machine Intelligence5 citationsFeb 20, 2025Turing.jl: A General-Purpose Probabilistic Programming LanguageGoogle DeepMind · ACM transactions on probabilistic machine learning.30 citationsFeb 15, 2025Navigating the Social Welfare Frontier: Portfolios for Multi-objective Reinforcement LearningAlphabet, Google DeepMind · arXiv (Cornell University)Feb 13, 2025On the Difficulty of Constructing a Robust and Publicly-Detectable WatermarkGoogle DeepMind · arXiv (Cornell University)1 citationFeb 7, 2025PartIR: Composing SPMD Partitioning Strategies for Machine LearningGoogle DeepMind · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS)4 citationsFeb 6, 2025Adapting to Evolving Adversaries with Regularized Continual Robust TrainingAlphabet, Google DeepMind · arXiv (Cornell University)Feb 6, 2025The Hidden Life of Tokens: Reducing Hallucination of Large Vision-Language Models via Visual Information SteeringGoogle DeepMind · arXiv (Cornell University)Feb 5, 2025Building Bridges between Regression, Clustering, and ClassificationGoogle DeepMind · arXiv (Cornell University)Feb 5, 2025Avoiding spurious sharpness minimization broadens applicability of SAMGoogle DeepMind · arXiv (Cornell University)Feb 4, 2025From Few to Many: Self-Improving Many-Shot Reasoners Through Iterative Optimization and GenerationAlphabet, Google DeepMind · arXiv (Cornell University)Feb 1, 2025SETS: Leveraging Self-Verification and Self-Correction for Improved Test-Time ScalingAlphabet, Google DeepMind · arXiv (Cornell University)1 citationJan 31, 2025Norm-Bounded Low-Rank AdaptationGoogle DeepMind · arXiv (Cornell University)Jan 31, 2025SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-trainingAlphabet, Google DeepMind · arXiv (Cornell University)7 citationsJan 28, 2025How should the advancement of large language models affect the practice of science?Google DeepMind · Proceedings of the National Academy of Sciences72 citationsJan 27, 2025Concurrent Learning with Aggregated States via Randomized Least Squares Value IterationGoogle DeepMind, Horizon Robotics · arXiv (Cornell University)Jan 23, 2025The Journey Matters: Average Parameter Count over Pre-training Unifies Sparse and Dense Scaling LawsAlphabet, Google DeepMind · arXiv (Cornell University)Jan 21, 2025MSTS: A Multimodal Safety Test Suite for Vision-Language ModelsGoogle DeepMind · arXiv (Cornell University)Jan 17, 2025Forget About It: Mitigating Forgetting Without Replay or Auxiliary ObjectivesAlphabet, Databricks, Google DeepMind · arXiv (Cornell University)Jan 9, 2025Reasoning-Enhanced Self-Training for Long-Form Personalized Text GenerationGoogle DeepMind · arXiv (Cornell University)Jan 7, 2025OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL CyclesAlphabet, Google DeepMind · neural information processing systems18 citationsJan 1, 2025Design2Code: Benchmarking Multimodal Code Generation for Automated Front-End EngineeringGoogle DeepMind, Microsoft · Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL)17 citationsJan 1, 2025