Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
Too Long to Watch, Too Short to Leave: Designing for Agent Micro-WaitsAlphabet · arXiv (Cornell University)Oct 19The Culture Within: Towards Mechanistic Understanding of Cultural Reasoning in LLMsLG AI Research · arXiv (Cornell University)Oct 19Scaling Synthetic Pre-training Data and Evaluation for Indic LLMsNVIDIA · arXiv (Cornell University)Oct 19Compiling Environment Classifiers to Contextual Modal TypesMeta · arXiv (Cornell University)Oct 19Fortress: Fork Resistant Transparency logs Relying on Social graph StructureMicrosoft · arXiv (Cornell University)Oct 19Secure Agentic Coding through Counterexample-Grounded FeedbackAlphabet · arXiv (Cornell University)Oct 19AfriGuard: Safety and Robust Refusal For African LanguagesCohere · arXiv (Cornell University)Oct 19LatComm: Reusable Context Compression for Latent Communication between Large Language ModelsNVIDIA · arXiv (Cornell University)Oct 19Breaking Diffusion: Parallel Reasoning with Draft RefinementMicrosoft · arXiv (Cornell University)Oct 19Focused Omni-Modal GUI AgentsAlibaba · arXiv (Cornell University)Oct 19Procedural Exposure Bias in Skill-Augmented Language Model AgentsAlibaba · arXiv (Cornell University)Oct 19RoAR-VAE: Improving Latent Diffusability through Robust PredictionMicrosoft · arXiv (Cornell University)Oct 19HarnessRL: Co-Evolving Policies with a Just-in-Time Training Harness for Open-Ended RLAlibaba · arXiv (Cornell University)Oct 19StepBench: Diagnosing Real-World Procedural Understanding in AI AgentsCohere, Microsoft · arXiv (Cornell University)Oct 19DeepSEA: A Multilingual Evaluation Benchmark for Southeast AsiaAlphabet · arXiv (Cornell University)Oct 19When Should AI Chime In? Benchmarking Timely Contribution of Proactive Agents in Multi-Party ChatAllen Institute for AI · arXiv (Cornell University)Oct 19VibeTraining-Bench: Data-Trainer-Harness Co-Design for Iterative System ImprovementAlibaba · arXiv (Cornell University)Oct 19Agent World Model in the Wild: Computer-Use Agent Environment SynthesisMicrosoft · arXiv (Cornell University)Oct 19Don't Rewrite the SID Map: Topology Preservation for Generative Retrieval in Dynamic E-Commerce SearchAlibaba · arXiv (Cornell University)Oct 19When Coding Agents Outpace Review: Evaluating PRD-Driven Development in Mature Enterprise RepositoriesAlibaba · arXiv (Cornell University)Oct 19WhatIf-Bench: Benchmarking Intervention-Aware Time Series ForecastingAlphabet · arXiv (Cornell University)Oct 19When the harness decides success: state that cannot be re-observedMicrosoft · arXiv (Cornell University)Oct 19RefCraft: Benchmarking Visual-Reference-Conditioned Agentic Generation through Web DevelopmentAlibaba · arXiv (Cornell University)Oct 19Neither Over- nor Under-Constrained: Evaluating LLM Personalization under a Constraint Qualification AuditTechnology Innovation Institute · arXiv (Cornell University)Oct 19OPDArena: Understanding Learning and Forgetting in On-Policy DistillationMicrosoft · arXiv (Cornell University)Oct 19