Empiricists of Computational Minds
Are you captivated by the rapid advancements in artificial intelligence? Do devour AI research papers and hack on with models in PyTorch? Do you contemplate how long the reign of transformers will last? If questions like these excite you, then this group is the perfect place for you.
Join us for an afternoon of stimulating discussions with like-minded individuals who are passionate about the empirical aspects of AI.
Why Attend?
Engage with Big Ideas:
The Future of Transformers: How much longer will Transformers reign supreme in the AI landscape? What could be the next paradigm shift in architectures?
Scaling Laws Enthusiasts: Do scaling laws not just interest but excite you? Let's discuss how they are shaping the way we understand model performance and capabilities.
Alignment with the Bitter Lesson: When evaluating new AI concepts, do you consider how they mesh with Rich Sutton's "The Bitter Lesson"? Let's explore the importance of leveraging computation over manual engineering.
Deep Dive into Core Questions:
Is Next-Token Prediction All We Need? Can predicting the next token lead us to general intelligence, or are there limitations we need to overcome?
The Role of Reward Systems: Is reward the fundamental component we should focus on? How does it compare to other learning signals in developing intelligent systems?
Embodied Intelligence: Is interacting with the real world through high-quality motor-sensory experiences essential for deep causal reasoning and robust world models?
Synthetic Data Solutions: Will synthetic data help us overcome the looming data scarcity? How can it be effectively integrated into current models?
Explore Mechanistic Interpretability:
If unraveling the inner workings of AI models is your passion, let's delve into mechanistic interpretability and discuss how it can make AI systems more transparent and trustworthy.
Discuss the Future of AI Development:
Balancing Data, Algorithms, and Compute: How will future progress in AI be distributed among better data, improved algorithms, and increased computational resources?
Who Should Attend?
AI Researchers and Practitioners: Share your insights and learn from peers tackling similar challenges.
Machine Learning Enthusiasts: Whether you're a professional or a hobbyist, your curiosity is welcome here.
Students and Academics: Connect with professionals in the field and explore opportunities for collaboration.