HyperBrowseComp: A Multilingual and Multimodal Stress Test for Web-Browsing Agents Paper • 2610.03574 • Published Oct 2 • 22
EditHero: A Benchmark for Long-Horizon Part-Level 3D Editing and Vibe Modeling Paper • 2610.02298 • Published Oct 1 • 15
RealCompanion: Benchmarking Human Understanding from Reasoning over Longitudinal Real-World Conversations Paper • 2610.01780 • Published Oct 1 • 39
Learning from Runtime Feedback through Failure-Bank Self-Evolution for Vision-Language-Action Models Paper • 2609.39820 • Published Sep 30 • 12
MinkowskiPE: Minkowski Positional Encoding for Spatiotemporal Perception Paper • 2609.33804 • Published Sep 27 • 7
When Does Dense Retrieval Need Asymmetric Geometry? A Bias-Variance Theory of Shared and Dual Projections Paper • 2609.32488 • Published Sep 26 • 9
EmoRES-TTS: Residual-Enhanced Vector Steering for Emotional Speech Generation Paper • 2609.38157 • Published Sep 29 • 15
OmniTaskonomy: When Does Visual Generation Improve Visual Understanding? Paper • 2609.38079 • Published Sep 29 • 41
Learning Beyond What You Sample: Off-Policy-Aware Cross-Model Trajectory Exchange for RLVR Paper • 2609.37868 • Published Sep 29 • 38
Think Before You Score: Thinking Reward Model for Visual Generation Paper • 2609.37372 • Published Sep 29 • 58
Raven: The Harness of Harnesses for Composable Agentic Intelligence Paper • 2609.33439 • Published Sep 27 • 100