First of all, thank you for your excellent work!
I am curious about the challenges involved in supporting ARC-3. ARC-3 is more targeted toward "interaction", where exploration and continual learning are required. In reinforcement learning, this seems to be a good testbed for evaluating how efficiently an agent can solve a task. Currently, ARC-1 and ARC-2 are supported. Is there any plan to support ARC-3?
First of all, thank you for your excellent work!
I am curious about the challenges involved in supporting ARC-3. ARC-3 is more targeted toward "interaction", where exploration and continual learning are required. In reinforcement learning, this seems to be a good testbed for evaluating how efficiently an agent can solve a task. Currently, ARC-1 and ARC-2 are supported. Is there any plan to support ARC-3?