YehuiTang
WizardTY
·
AI & ML interests
None yet
Recent Activity
posted an update 8 days ago
If you want a hareness tuned for your scenario, try Gear. We just open-sourced a RSI infra for adapting agents to real-world tasks. On Automation bench, Gear-tuned DSH + Luna beats Astra + Codex on objective completion rate (84% vs 84%), with 95% lower API cost.
try Gear: https://github.com/rsi-gear/gear authored a paper over 1 year ago
DualTHOR: A Dual-Arm Humanoid Simulation Platform for Contingency-Aware
PlanningOrganizations
None yet