[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"sanity-MplAdE7xj1VGTGNTavJlUqUS5Dv7ZgPyzCGTFimLYPE":3},{"data":4,"sourceMap":-1},{"info":5,"latest":15,"pinned":63},{"_createdAt":6,"_id":7,"_rev":8,"_type":9,"_updatedAt":6,"descriptionText":10,"slug":11,"title":14},"2024-04-04T14:35:24Z","d0a9b59b-c92e-48f3-84ea-4a9d1093b4e3","l85YFb2NNiT2L5SEvOgTDc","blogTag",null,{"_type":12,"current":13},"slug","mixture-of-experts","mixture of experts",[16],{"_id":17,"author":18,"commentCount":30,"comments":31,"excerpt":32,"featureTag":10,"image":33,"publishedAt":37,"slug":38,"sponsored":10,"tags":40,"title":62},"bf6cc061-dcb9-41cd-bfa4-6d15b2421dd9",[19],{"_id":20,"avatar":21,"name":26,"role":27,"slug":28},"12169ed9-78d5-4f82-adb4-e956348db389",{"_type":22,"asset":23},"image",{"_ref":24,"_type":25},"image-7c75ba835850e53201ea979015e8b8f449e126da-400x400-png","reference","Cameron R. Wolfe, PhD","Director of AI at Rebuy",{"_type":12,"current":29},"cameron-r-wolfe-phd",0,true,"This new LLM technique has started improving the results of models without additional training. ",{"_type":22,"asset":34,"attribution":36},{"_ref":35,"_type":25},"image-806103733e75354ebebd71aa8baf817451aec6e1-1280x672-jpg","Alexandra Francis","2024-04-04T14:31:11.098Z",{"_type":12,"current":39},"how-do-mixture-of-experts-layers-affect-transformer-models",[41,48,52,54,58],{"_createdAt":42,"_id":43,"_rev":44,"_type":9,"_updatedAt":42,"slug":45,"title":47},"2023-05-23T16:43:21Z","wp-tagcat-generative-ai","9HpbCsT2tq0xwozQfkc4ih",{"current":46},"generative-ai","generative AI",{"_createdAt":42,"_id":49,"_rev":44,"_type":9,"_updatedAt":42,"slug":50,"title":51},"wp-tagcat-llm",{"current":51},"llm",{"_createdAt":6,"_id":7,"_rev":8,"_type":9,"_updatedAt":6,"slug":53,"title":14},{"_type":12,"current":13},{"_createdAt":42,"_id":55,"_rev":44,"_type":9,"_updatedAt":42,"slug":56,"title":57},"wp-tagcat-se-tech",{"current":57},"se-tech",{"_createdAt":42,"_id":59,"_rev":44,"_type":9,"_updatedAt":42,"slug":60,"title":61},"wp-tagcat-se-stackoverflow",{"current":61},"se-stackoverflow","How do mixture-of-experts layers affect transformer models?",[]]