Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
Efficient Adaptive Federated OptimizationAlphabet · neural information processing systems1 citationJan 1, 2025OmniBench: Towards The Future of Universal Omni-Language ModelsAlibaba, Google DeepMind · neural information processing systems1 citationJan 1, 2025EUGens: Efficient, Unified and General Dense LayersAlphabet, Google DeepMind · neural information processing systems1 citationJan 1, 20253D-Prover: Diversity Driven Theorem Proving With Determinantal Point ProcessesGoogle DeepMind · neural information processing systems1 citationJan 1, 2025Benign Overfitting in Single-Head AttentionAlphabet · neural information processing systems1 citationJan 1, 2025DUO: No Compromise to Accuracy DegradationAlphabet · neural information processing systemsJan 1, 2025Understanding Prompt Tuning and In-Context Learning via Meta-LearningGoogle DeepMind · neural information processing systemsJan 1, 2025Principled Model Routing for Unknown Mixtures of Source DomainsAlphabet · neural information processing systemsJan 1, 2025Learning Interestingness in Automated Mathematical Theory FormationGoogle DeepMind · neural information processing systemsJan 1, 2025Depth-Width Tradeoffs for Transformers on Graph TasksAlphabet · neural information processing systemsJan 1, 2025Progressive Data Dropout: An Embarrassingly Simple Approach to Train FasterGoogle DeepMind · neural information processing systemsJan 1, 2025Enhancing Tactile-based Reinforcement Learning for Robotic ControlGoogle DeepMind · neural information processing systemsJan 1, 2025Chirality in Action: Time-Aware Video Representation Learning by Latent StraighteningGoogle DeepMind · neural information processing systemsJan 1, 2025Thoughts Are All Over the Place: On the Underthinking of Long Reasoning ModelsAlphabet · neural information processing systemsJan 1, 2025Length Generalization via Auxiliary TasksAlphabet · neural information processing systemsJan 1, 2025Learn and Ensemble Bridge Adapters for Multi-domain Task Incremental LearningAlphabet · neural information processing systemsJan 1, 2025Videos are Sample-Efficient Supervisions: Behavior Cloning from Videos via Latent RepresentationsAlphabet · neural information processing systemsJan 1, 2025Explaining and Mitigating Crosslingual Tokenizer InequitiesGoogle DeepMind · neural information processing systemsJan 1, 2025On the Emergence of Linear Analogies in Word EmbeddingsGoogle DeepMind · neural information processing systemsJan 1, 2025From Style to Facts: Mapping the Boundaries of Knowledge Injection with FinetuningAlphabet · neural information processing systemsJan 1, 2025Tapered Off-Policy REINFORCE - Stable and efficient reinforcement learning for large language modelsGoogle DeepMind, Microsoft · neural information processing systemsJan 1, 2025Spark Transformer: Reactivating Sparsity in Transformer FFN and AttentionAlphabet, Google DeepMind · neural information processing systemsJan 1, 2025On Union-Closedness of Language GenerationAlphabet · neural information processing systemsJan 1, 2025Massive Sound Embedding Benchmark (MSEB)Alphabet · neural information processing systemsJan 1, 2025Democratizing Clinical Risk Prediction with Cross-Cohort Cross-Modal Knowledge TransferAlphabet · neural information processing systemsJan 1, 2025