3 ms·
Do you see these trajectories being used to fine tune a model automatically in some way rather then just replay, that way similar workflows might be improved to
by dbish 1y ago
Do you see these trajectories being used to fine tune a model automatically in some way rather then just replay, that way similar workflows might be improved too?
- edunteman 1y agoI believe explicit trajectories for learned behavior are significantly easier for humans to grok and debug, in contrast to reinforcement learning methods like deep Q-learning, so avoiding the use of models is ideal, but I imagine they'll have their place. For what that may look like, I'll reuse a brainstorm on this topic that a friend gave me recently: "Instead of relying on an LLM to understand where to click, the click area itself is the token. And the click is a token and the objective is a token and the output is whatever. Such that, click paths aren't "stored", they're embedded within the training of the LAM/LLM" Whatever it ends up looking like, as long as it gets the job done and remains debuggable and extensible enough to not immediately eject a user once they hit any level of complexity, I'd be happy for it to be a part of Muscle Mem.