https://phillipi.github.io/prh The Platonic Representation HypothesisMinyoung Huh, Brian Cheung, Tongzhou Wang, Phillip IsolaMITPosition Paper in ICML 2024 “Neural networks, trained with different objectives on different data and modalities, are converging to
What I Read: Chain-of-Thought Reasoning
https://www.quantamagazine.org/how-chain-of-thought-reasoning-helps-neural-networks-compute-20240321 How Chain-of-Thought Reasoning Helps Neural Networks ComputeBen Brubaker3/21/24 11:15 AM “Large language models do better at solving problems when they show their work. Researchers are beginning to understand why.”
What I Read: Artificial, Biological Intelligence
https://medium.com/@begus.gasper/artificial-and-biological-intelligence-humans-animals-and-machines-142bc3c4b304 Artificial and Biological Intelligence: Humans, Animals, and MachinesGasper BegusSep 19, 2023 “…a highly promising direction in AI research is to use artificial intelligence to better understand biological intelligence, and
What I Read: Chatbots Understand Text
https://www.quantamagazine.org/new-theory-suggests-chatbots-can-understand-text-20240122/ New Theory Suggests Chatbots Can Understand TextAnil Ananthaswamy1/22/24 “Far from being “stochastic parrots,” the biggest large language models seem to learn enough skills to understand the words they’re processing.”
What I Read: AI System Beats Chess Puzzles
https://www.quantamagazine.org/google-deepmind-trains-artificial-brainstorming-in-chess-ai-20231115/ AI System Beats Chess Puzzles With ‘Artificial Brainstorming’Stephen OrnesNovember 15, 2023 “By bringing together disparate approaches, machines can reach a new level of creative problem-solving.”