Yu Su Intelligence Continual Learning Expertise

Summary

Scheduling a meeting is not finding a shared slot on everyone's calendar. It is a constraint optimization over authority, priority, and ... A presentation from the 2022 Artificial Talk Title: Computer Use: Modern Moravec's Paradox Associate Professor · Training a model on new data or tasks over time without forgetting previously learned At supermemory, we are doubling down and pushing the frontier for memory and in-context We have models that pass the bar exam and write functional code in seconds. But if you actually use them for real work, you ... Lex Fridman Podcast full episode: youtube.com/watch?v=EV7WhVT270Q Thank you for listening ❤ our ... So some of you might have noticed that these issues are deeply related to existing research directions in Every leaderboard you have seen was built by asking a model to do one task, wiping its memory, and asking it another. ContinualAI Seminar: "NISPA: Neuro-Inspired Stability-Plasticity Adaptation for