Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
ShredBench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLMs in Document ReconstructionAlibaba · arXiv (Cornell University)Apr 26The Network Structure of MathlibAlphabet · arXiv (Cornell University)Apr 26ShredBench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLMs in Document ReconstructionAlibaba · arXiv (Cornell University)Apr 26SFT-then-RL Outperforms Mixed-Policy Methods for LLM ReasoningAllen Institute for AI · arXiv (Cornell University)Apr 26DRACULA: Hunting for the Actions Users Want Deep Research Agents to ExecuteAllen Institute for AI · arXiv (Cornell University)Apr 26COMET: A Framework for Modeling Compound Operation Dataflows with Explicit Collectivesd-Matrix · IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)Apr 26Probabilistic analysis of dual decomposition on two-stage stochastic integer programsMicrosoft · arXiv (Cornell University)Apr 25Disentangling Representations from Search Behaviors for Recommendation via Counterfactual LearningAlibaba · ACM Transactions on Information SystemsApr 25REPS: Recycled Entropy Packet Spraying for Adaptive Load Balancing and Failure MitigationMicrosoft · European Conference on Computer Systems12 citationsApr 24GAIR: Location-aware self-supervised contrastive pre-training with geo-aligned implicit representationsAlphabet · ISPRS Journal of Photogrammetry and Remote Sensing4 citationsApr 24How Do AI Agents Spend Your Money? Analyzing and Predicting Token Consumption in Agentic Coding TasksAlphabet, Google DeepMind, Microsoft · arXiv (Cornell University)3 citationsApr 24Canopy: Property-Driven Learning for Congestion ControlAlphabet · European Conference on Computer Systems3 citationsApr 24Bridging the GPU Utilization Gap: Predictive Multi-Dimensional Resource Scheduling for AI WorkloadsAlibaba · European Conference on Computer Systems1 citationApr 24AIMS: Cost-Efficient LLM-Based Agent Deployment in Hybrid Cloud-Edge EnvironmentsMicrosoft · European Conference on Computer Systems1 citationApr 24GPT-Eval: A Unified Evaluation Framework for Tools and Text ResponsesMicrosoft · Cognitive Models and Artificial Intelligence Conference (AICCONF)Apr 24CRAFT: Clustered Regression for Adaptive Filtering of Training dataAlphabet · arXiv (Cornell University)Apr 24How LLMs Detect and Correct Their Own Errors: The Role of Internal Confidence SignalsAlphabet · arXiv (Cornell University)Apr 24CRAFT: Clustered Regression for Adaptive Filtering of Training dataAlphabet · arXiv (Cornell University)Apr 24PASS: A Power Adaptive Storage ServerDatabricks · European Conference on Computer SystemsApr 24Avicenna: Masking Slowdowns in Replicated State Machines with Counterfactual EvaluationDatabricks, Microsoft · European Conference on Computer SystemsApr 24LightDSA: Enabling Efficient DSA Through Hardware-Aware Transparent OptimizationAlibaba · European Conference on Computer SystemsApr 24Concord: Learning Network Configuration ContractsMicrosoft · European Conference on Computer SystemsApr 24DROPS: Managing Serverless Resource Pools in Microsoft Azure FunctionsMicrosoft · European Conference on Computer SystemsApr 24PiLLM: Resource-Efficient LLM Inference Using Workload PredictionSenseTime · European Conference on Computer SystemsApr 24Scheduling Cloud Block Storage Proactively and Reactively with OmarAlibaba · European Conference on Computer SystemsApr 24