Template Matching Insights
Timothy discusses the complexities of robust prediction in transformers, emphasizing the need for flexible rule systems and template negotiation. He highlights the distinction between description and explanation, using the example of the sun's movement to illustrate how different interpretations can arise from the same observations. Keith adds depth to the conversation by touching on the entanglement of data and the challenges of disentangling original semantics in syntactic processing, referencing the Chinese room argument to underscore the limitations of understanding.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Is ChatGPT an N-gram model on steroids?
Related Questions