Thanks for the work, this is highly engaging and relevant to current research trends, given the rapid proliferation of applying Large Language Models (LLMs).
I was wondering about the possibility of adding a new property as having a sequence of tasks such as a continual lifelong learning framework.
Consider, for instance, the process of preparing a meal like a mashed potato, which involves a series of interconnected tasks such as growing potatoes, boiling water, cooking the potatoes, setting the table (finding non-living things), and serving the dish.
I thought introducing this sequential aspect might bring the environment much closer to real-world scenarios, creating a more realistic setup for continual learning works.
Thanks for the work, this is highly engaging and relevant to current research trends, given the rapid proliferation of applying Large Language Models (LLMs). I was wondering about the possibility of adding a new property as having a sequence of tasks such as a continual lifelong learning framework. Consider, for instance, the process of preparing a meal like a mashed potato, which involves a series of interconnected tasks such as growing potatoes, boiling water, cooking the potatoes, setting the table (finding non-living things), and serving the dish. I thought introducing this sequential aspect might bring the environment much closer to real-world scenarios, creating a more realistic setup for continual learning works.