Meta’s breakthrough: Byte Latent TransformersPatches Scale Better Than TokensDec 17, 2024Dec 17, 2024
Representation Finetuning (ReFT) for Breakthrough PerformanceDramatically improve model performance with fewer resourcesDec 14, 2024Dec 14, 2024
EMOQ (v4 update 28/12)A Symbolic Framework for Integrated Cognitive-Emotional Human-AI CommunicationDec 11, 2024Dec 11, 2024
Physics in Next-Token Prediction: Physical Basis of Machine IntelligenceIntroduction: Physics and AI IntelligenceNov 10, 2024Nov 10, 2024
Probabilistic Inference in Language Models via Twisted Sequential Monte CarloOct 26, 2024Oct 26, 2024
The mathematical foundations of transformersby fhsp ( fernandohenriquesp@gmail.com )Aug 20, 2024Aug 20, 2024
