Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
SeKV: Resolution-Adaptive KV Cache with Hierarchical Semantic Memory for Long-Context LLM InferenceMicrosoft · arXiv (Cornell University)Jun 30Introduction to Stochastic Differential Equations for Generative Machine Learning: A Variational PerspectiveGoogle DeepMind · arXiv (Cornell University)Jun 30Bridging the Gap Between Latent and Explicit Reasoning with Looped TransformersMicrosoft · arXiv (Cornell University)Jun 30Reinforcement Learning with Metacognitive Feedback Elicits Faithful Uncertainty Expression in LLMsAlphabet · arXiv (Cornell University)Jun 30CDR-Bench: Evaluating Faithful Execution of Compositional, Order-Sensitive Data Refinement RecipesAlibaba · arXiv (Cornell University)Jun 30SmartDispatch: Dynamic Substitution of NumPy-Style APIs on Heterogeneous CPU-GPU SystemsMeta · Proceedings of the ACM on software engineering.Jun 30Investigating the Effect of Class Imbalance on the Performance of Self-Admitted Technical Debt Prediction ModelsAlibaba · International Journal of Software Engineering and Knowledge EngineeringJun 30TSGuard: Automated User-Centric Incident Diagnosis for AI Workloads in the CloudMicrosoft · Proceedings of the ACM on software engineering.Jun 30Real-time 3D Visualization of Radiance Fields on Light Field DisplaysNVIDIA · Proceedings of the ACM on Computer Graphics and Interactive Techniques1 citationJun 29Indi-RomCoM: Code-Mixed Benchmark for Evaluating LLMs on Romanized Indic-English InstructionsAlphabet · arXiv (Cornell University)Jun 29Understanding and Evaluating Claw-like Agent Security Through a Computer-Systems LensAlphabet · arXiv (Cornell University)Jun 29Preserving Speech-to-Text LLM Capabilities in Speech-to-Speech GenerationMicrosoft · arXiv (Cornell University)Jun 29LatentRevise: Learning from Zero-Hit ReasoningMicrosoft · arXiv (Cornell University)Jun 29ConCent: Contact-Centric Real-to-Sim-to-Real Learning from One DemonstrationMicrosoft · arXiv (Cornell University)Jun 29Dynamo: Dynamic Skill-Tool Evolution for Vision-Language AgentsAlibaba · arXiv (Cornell University)Jun 29Mandol: An Agglomerative Agent Memory System for Long-Term ConversationsMicrosoft · arXiv (Cornell University)Jun 29Compatibility-Guided Neighbor Selection for ReSTIRNVIDIA · Proceedings of the ACM on Computer Graphics and Interactive TechniquesJun 29Warnock: Harnessing GPU Geometry Amplification for Vector GraphicsNVIDIA · Proceedings of the ACM on Computer Graphics and Interactive TechniquesJun 29Scalable Training and Rendering of 3D Gaussians for Large Scale Scientific DataNVIDIA · Proceedings of the ACM on Computer Graphics and Interactive TechniquesJun 29Real-time Path Tracing of Massive Dynamic FoliageNVIDIA · Proceedings of the ACM on Computer Graphics and Interactive TechniquesJun 29Quasiparticle-induced decoherence of a driven superconducting qubitAlphabet · Physical Review AppliedJun 29A Multi-Dataset Benchmark for Evaluating LLM Agents in Microservice Failure DiagnosisAlibaba · arXiv (Cornell University)Jun 28The Platonic Defense: Backdoor Defense for Self-Supervised Encoders in the Era of Large Scale Pre-trainingAlibaba · arXiv (Cornell University)Jun 28Note on Finite-Automata Bernoulli Factories for Rational FunctionsAlphabet · arXiv (Cornell University)Jun 28TF-MoE: Time-Frequency Mixture-of-Experts for Efficient Speech SeparationMicrosoft · arXiv (Cornell University)Jun 28