Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
EvidenT: An Evidence-Preserving Framework for Iterative System-Level Package RepairMicrosoft · Proceedings of the ACM on software engineering.Oct 1Automated Modernization of Machine Learning Engineering Notebooks for ReproducibilityAlphabet · Proceedings of the ACM on software engineering.Oct 1Unitary fault-tolerant encoding of Pauli states in surface codesNVIDIA · npj Quantum InformationOct 1Automatically Generating ML Compiler Backends from Tensor Accelerator ISA DescriptionsNVIDIA · Proceedings of the ACM on Programming LanguagesOct 1Bayesian Nonparametric Risk Assessment in Developmental Toxicity Studies With Ordinal ResponsesAlphabet · EnvironmetricsOct 1Polynomial-Time Cryptanalytic Extraction of Deep Neural Networks in the Hard-Label SettingGoogle DeepMind, Technology Innovation Institute · Journal of CryptologyOct 1Function-preserving watermarking of AI-generated proteinsGoogle DeepMind · Nature2 citationsSep 30ContextFidelity-Bench: A Three-Paradigm Framework for Evaluating Multi-Turn Hallucination Behavior in Language ModelsAlphabet · BenchCouncil Transactions on Benchmarks Standards and EvaluationsSep 30Quantum Algorithms for OPI Variants Beyond Locality and Classical DecodabilityAlphabet · arXiv (Cornell University)Sep 30Decoupling of the QAOA into independent spin-boson systems and high-depth performance on pure and mixed spin glassesAlphabet · arXiv (Cornell University)Sep 30Mixture of Decoders for Diverse Dialog Response GenerationMicrosoft · arXiv (Cornell University)Sep 30MatrixReward: Reward from Rubric Matrix for Open-Ended GenerationAlibaba · arXiv (Cornell University)Sep 30The Life Cycle of a Massive Activation: Stochastic Birth, Weight-Decay-Driven Growth, and Competitive ConsolidationMistral AI · arXiv (Cornell University)Sep 30MegaFlux: Skew-Resilient MoE Megakernels via Pipelined Expert ReplicationNVIDIA · arXiv (Cornell University)Sep 30Where's Waldo? Query-language Preference under Cross-lingual Knowledge DisparitiesMicrosoft · arXiv (Cornell University)Sep 30Gumbel Straight Flow: Distilling Autoregressive Models into One-step Flow MapsGoogle DeepMind · arXiv (Cornell University)Sep 30Herschel: Continuous Optimization of Production LLM Inference through On-Demand ProfilingAlibaba · arXiv (Cornell University)Sep 30A Fourier-Label Information-Loss Barrier for Dihedral Coset AlgorithmsAlphabet · arXiv (Cornell University)Sep 30SCIC: Scope- and Codebook-Aware Instruction Conditioning for Speaker-Adapted Expressive TTSAlibaba · arXiv (Cornell University)Sep 30Capture the lifecycle: KV Cache management in ReAct Agents with KVTetherAlibaba · arXiv (Cornell University)Sep 30GFD-OPD: Guidance-Folded On-Policy Distillation of Diffusion Models Across ScalesZhipu AI · arXiv (Cornell University)Sep 30Covert Assistance: Helpful LLM Agents Evade Oversight in Multi-Agent SystemsNVIDIA · arXiv (Cornell University)Sep 30Multimodal Flow: Unified Flow Modeling of Language and Vision in Embedding SpacesHorizon Robotics · arXiv (Cornell University)Sep 30ID Balancing: Stable Training of Extremely Sparse MoE via PID-Based Load ControlAlibaba · arXiv (Cornell University)Sep 30Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical PriorHugging Face · arXiv (Cornell University)Sep 30