George SanchezUnderstanding Large Language Models with Sparse Autoencoders: Potential and ChallengesIn early June, OpenAI published a paper detailing training a 16 million latent autoencoder on GPT-4 activations using 40 billion tokens…Jul 5Jul 5
George SanchezLinguistic Might: Shaping Beliefs Using Large Language ModelsThe profound impact of language on beliefs is evident, with words wielding the power to shape perspectives and ideologies. Researchers such…Feb 28Feb 28
George SanchezEffective Communication Strategies: Lessons from Nursery RhymesEven a nursery rhyme like “Baa, Baa, Black Sheep” can offer valuable insights into effective communication. This discussion aims to draw…Feb 22Feb 22
George SanchezEmergent Abilities in Large Language ModelsThe popularity of large language models (LLMs) is marked by a phenomenon that captures the imagination of AI practitioners — the emergence…Jan 25Jan 25
George SanchezDiffusion in Generative AI for NLPA generative model strives to discern the intricacies of data distribution from observed samples, empowering itself to draw samples at will…Dec 22, 2023Dec 22, 2023
George SanchezThe Significance of Language Patterns in Leveraging Large Language Models (LLM)Language, with its inherent pattern attribute, is the backbone for the remarkable capabilities of Large Language Models (LLMs). These…Nov 17, 2023Nov 17, 2023
George SanchezReimagining AI: Going Beyond Getting Things RightIn today’s world of artificial intelligence and neural networks, some people are questioning what intelligence means for machines and…Oct 21, 2023Oct 21, 2023
George SanchezTrust me. I am an Expert!Organizations are demanding innovation and paradigm shifts. In my work as a research scientist, we are often given challenges requiring…Dec 22, 2022Dec 22, 2022