UC Berkeley · 11 records

dateidtitleauthorsvenue↑
2026-11-022610.37651GEPA-Curator: Curated Learning for Self Improving AgentsZack Yu et al.arXiv6
2026-10-222610.28940StyleStream 2.0: Fast and Controllable Streaming Voice Style ConversionYisi Liu et al.arXiv16
2026-10-152610.18755PreciseCoder: Faithful and Generalizable Debugging Beyond Unit-Test RewardsYunkai Zhan et al.arXiv9
2026-10-052610.03710EyeRobot 2.0: Active Gaze for Precise Manipulation without Wrist CamerasKush Hari et al.arXiv7
2026-09-302609.38079OmniTaskonomy: When Does Visual Generation Improve Visual Understanding?Jiaxin Ge et al.arXiv45
2026-09-282609.28660Morphometric Imitation: From Morphology and Contact Aware Hand Retargeting to Sim-to-Real Visuomotor PolicyTara Sadjadpour et al.arXiv12
2026-09-172609.19144A Zeroth-Order Paradigm for LLM Preference AlignmentPeter Chen et al.arXiv12
2026-07-072607.05369GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation TasksKaiyuan Chen et al.arXiv5
2026-06-192606.19419Playful Agentic Robot LearningJunyi Zhang et al.arXiv49
2026-06-032606.05405Agents' Last ExamYiyou Sun et al.arXiv273
2026-06-012605.26112From Model Scaling to System Scaling: Scaling the Harness in Agentic AIShangding GuarXiv5