Explorative Modeling – Unlocking a Third Pretraining Axis and End-to-End Generation
What in the world is Explorative Modeling?
blank
Alexi Gladstone's official website.
Latest posts
Training for Marathons by Sprinting—Why Exposure Bias is Evil
Autoregressive and diffusion models are trained on a single step but inferenced on thousands. Why does this work at all?
Nurturing Superintelligence
Why Will AI Be Superintelligent?
The Future of Benchmarking and the Meaning of Life
What's a benchmarking paradigm that'll scale to human-level intelligence? What's the meaning of life? How are these related?
The Future of Multimodal Does not Involve Text
What multimodal paradigm will win out?
Energy-Based Transformers are Scalable Learners and Thinkers
What are Energy-Based Transformers and why should I care about them?
The Different Components of Intelligence
What are the different components of intelligence?
How Can We Improve Governments?
How can we change motivations to improve governments?
The Neuroplasticity Hypothesis
What do many of the most impactful deep learning architectural components have in common?
Does Biological Intelligence Have Any Advantages Over Digital Intelligence?
How does creativity come about?