Where am I?
What have I seen?
What is probably behind that doorway?
Which obstacles are real vs perception artifacts?
Which parts of the map are uncertain?
What is my localization confidence?
What actions reduce uncertainty?
What actions are unsafe?
What does the human instruction mean in this scene?
When should I continue, retreat, hover, or ask?
What have I seen?
What is probably behind that doorway?
Which obstacles are real vs perception artifacts?
Which parts of the map are uncertain?
What is my localization confidence?
What actions reduce uncertainty?
What actions are unsafe?
What does the human instruction mean in this scene?
When should I continue, retreat, hover, or ask?
💡 Remember Box
https://huggingface.co/UofTCSSLab/Maia3-79M
Maia Chess
Introducing Maia: a Human-Like Chess Engine
We develop and introduce Maia, a machine learning model that captures human style in chess. Maia is trained on millions of games, and tries to predict the human move played in each position seen.
DiffusionBlocks: Training Neural Networks One Block at a Time
https://pub.sakana.ai/diffusionblocks/
https://pub.sakana.ai/diffusionblocks/
Sakana AI
DiffusionBlocks: Training Neural Networks One Block at a Time
DiffusionBlocks converts residual networks into independently trainable blocks via a diffusion interpretation, cutting training memory by B× while matching end-to-end performance.