ssillyssadass@lemmy.world to Lemmy Shitpost@lemmy.world · 2 months agoNeed a keyboard with a dedicated "slop" buttonlemmy.worldimagemessage-square189fedilinkarrow-up1622arrow-down1124
arrow-up1498arrow-down1imageNeed a keyboard with a dedicated "slop" buttonlemmy.worldssillyssadass@lemmy.world to Lemmy Shitpost@lemmy.world · 2 months agomessage-square189fedilink
minus-squareBonsoir@lemmy.calinkfedilinkarrow-up3arrow-down1·2 months agoThe precomputed matrices are definitely a neural network. GPT literally stands for generative pre-trained transformer, a transformer being a well known neural network architecture.
minus-squareBonsoir@lemmy.calinkfedilinkarrow-up1arrow-down1·2 months agohttps://arxiv.org/pdf/1706.03762
minus-squareBonsoir@lemmy.calinkfedilinkarrow-up1arrow-down1·2 months agoThen explain to me how the neural networks behind LLMs are not mimicking neurons. Because that’s the basic idea behind all of deep learning.
deleted by creator
The precomputed matrices are definitely a neural network. GPT literally stands for generative pre-trained transformer, a transformer being a well known neural network architecture.
deleted by creator
https://arxiv.org/pdf/1706.03762
deleted by creator
Then explain to me how the neural networks behind LLMs are not mimicking neurons. Because that’s the basic idea behind all of deep learning.