benchmark · first seen 27 Oct 2026
real-time interaction benchmarking for vision-language agents
Also: Gym-Anything-RT
Evaluation of AI agents in visual minigames requiring real-time perception and action under an evolving environment state.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Gym-Anything-RT: Scaling Real-Time Environments for Vision-Language AgentsPapers · 27 Oct 2026 · Pranjal Aggarwal, Sean Welleck
Latest sources
- Gym-Anything-RT: Scaling Real-Time Environments for Vision-Language AgentsPapers · 27 Oct 2026 · Pranjal Aggarwal, Sean Welleck