https://mewelch.substack.com/p/putting-the-human-touch-on-llms Putting the human touch on LLMsMolly WelchMar 30 “Techniques like RLHF help align large language models with people’s values and preferences. Is that a good thing?”
What I Read: GPT, Ranking
https://messyprogress.substack.com/p/gpt-is-rather-good-at-feed-ranking GPT is Rather Good at Feed RankingRob EnnalsMar 7 “If ranking is as easy as saying what should rank highly, then lots of interesting things happen.”
What I Read: Abilities Emerging From AI
https://www.quantamagazine.org/the-unpredictable-abilities-emerging-from-large-ai-models-20230316/ The Unpredictable Abilities Emerging From Large AI Models “Large language models like ChatGPT are now big enough that they’ve started to display startling, unpredictable behaviors.”
What I Read: Teach Computers Math
https://www.quantamagazine.org/to-teach-computers-math-researchers-merge-ai-approaches-20230215/ To Teach Computers Math, Researchers Merge AI ApproachesKevin HartnettFebruary 15, 2023 “Large language models still struggle with basic reasoning tasks. Two new papers that apply machine learning to math
What I Read: Machines Learn, Teach Basics
https://www.quantamagazine.org/machines-learn-better-if-we-teach-them-the-basics-20230201/ Machines Learn Better if We Teach Them the BasicsMax G. LevyFebruary 1, 2023 “A wave of research improves reinforcement learning algorithms by pre-training them as if they were human.”