LoveOxygenProducers@lemmy.worldEnglish · 1 year ago[R] Unraveling the Mysteries: Why is AdamW Often Superior to Adam+L2 in Practice?plus-squaremessage-squaremessage-square3fedilinkarrow-up118arrow-down11
arrow-up117arrow-down1message-square[R] Unraveling the Mysteries: Why is AdamW Often Superior to Adam+L2 in Practice?plus-squareLoveOxygenProducers@lemmy.worldEnglish · 1 year agomessage-square3fedilink
abhi9u@lemmy.worldEnglish · 1 year agoAn Analysis of DeepMind's 'Language Modeling Is Compression' Paperplus-squarecodeconfessions.substack.comexternal-linkmessage-square0fedilinkarrow-up110arrow-down11
arrow-up19arrow-down1external-linkAn Analysis of DeepMind's 'Language Modeling Is Compression' Paperplus-squarecodeconfessions.substack.comabhi9u@lemmy.worldEnglish · 1 year agomessage-square0fedilink
kromem@lemmy.worldEnglish · 1 year agoThe Physical Process That Powers a New Type of Generative AIplus-squarewww.quantamagazine.orgexternal-linkmessage-square0fedilinkarrow-up111arrow-down10
arrow-up111arrow-down1external-linkThe Physical Process That Powers a New Type of Generative AIplus-squarewww.quantamagazine.orgkromem@lemmy.worldEnglish · 1 year agomessage-square0fedilink
kromem@lemmy.worldEnglish · 1 year agoMachine-learning system based on light could yield more powerful, efficient large language modelsplus-squarenews.mit.eduexternal-linkmessage-square0fedilinkarrow-up19arrow-down11
arrow-up18arrow-down1external-linkMachine-learning system based on light could yield more powerful, efficient large language modelsplus-squarenews.mit.edukromem@lemmy.worldEnglish · 1 year agomessage-square0fedilink
AlmightySnoo 🐢🇮🇱🇺🇦@lemmy.worldEnglish · 1 year agoRisky Giant Steps Can Solve Optimization Problems Fasterplus-squarewww.quantamagazine.orgexternal-linkmessage-square0fedilinkarrow-up14arrow-down10
arrow-up14arrow-down1external-linkRisky Giant Steps Can Solve Optimization Problems Fasterplus-squarewww.quantamagazine.orgAlmightySnoo 🐢🇮🇱🇺🇦@lemmy.worldEnglish · 1 year agomessage-square0fedilink
Bluetreefrog@lemmy.worldEnglish · 1 year agoRecommendations for a context aware text classifierplus-squaremessage-squaremessage-square3fedilinkarrow-up110arrow-down10
arrow-up110arrow-down1message-squareRecommendations for a context aware text classifierplus-squareBluetreefrog@lemmy.worldEnglish · 1 year agomessage-square3fedilink
Bluetreefrog@lemmy.worldEnglish · 1 year agoAgencies for ML jobs in Eastern Australiaplus-squaremessage-squaremessage-square0fedilinkarrow-up16arrow-down11
arrow-up15arrow-down1message-squareAgencies for ML jobs in Eastern Australiaplus-squareBluetreefrog@lemmy.worldEnglish · 1 year agomessage-square0fedilink
Jure Repinc@lemmy.mlEnglish · 1 year ago“AI” Hurts Consumers and Workers -- and Isn’t Intelligentplus-squaretechpolicy.pressexternal-linkmessage-square0fedilinkarrow-up14arrow-down14
arrow-up10arrow-down1external-link“AI” Hurts Consumers and Workers -- and Isn’t Intelligentplus-squaretechpolicy.pressJure Repinc@lemmy.mlEnglish · 1 year agomessage-square0fedilink
soyagi@yiffit.netEnglish · 1 year agoNew AI systems collide with copyright lawplus-squarewww.bbc.co.ukexternal-linkmessage-square0fedilinkarrow-up18arrow-down10
arrow-up18arrow-down1external-linkNew AI systems collide with copyright lawplus-squarewww.bbc.co.uksoyagi@yiffit.netEnglish · 1 year agomessage-square0fedilink
f4hy@lemmy.worldEnglish · 1 year agoLooking for resources on music generationplus-squaremessage-squaremessage-square2fedilinkarrow-up18arrow-down10
arrow-up18arrow-down1message-squareLooking for resources on music generationplus-squaref4hy@lemmy.worldEnglish · 1 year agomessage-square2fedilink
LordShrek@lemmy.worldEnglish · 1 year agoDiscussion of llama source codeplus-squaremessage-squaremessage-square1fedilinkarrow-up18arrow-down11
arrow-up17arrow-down1message-squareDiscussion of llama source codeplus-squareLordShrek@lemmy.worldEnglish · 1 year agomessage-square1fedilink
realz@lemmy.worldEnglish · 1 year agoWhat tools/libraries do you for MLOps?plus-squaremessage-squaremessage-square1fedilinkarrow-up18arrow-down11
arrow-up17arrow-down1message-squareWhat tools/libraries do you for MLOps?plus-squarerealz@lemmy.worldEnglish · 1 year agomessage-square1fedilink
EthicalAI@lemmy.worldEnglish · 1 year ago3D-LLM Injecting the 3D World into Large Language Modelsvis-www.cs.umass.eduexternal-linkmessage-square0fedilinkarrow-up18arrow-down11
arrow-up17arrow-down1external-link3D-LLM Injecting the 3D World into Large Language Modelsvis-www.cs.umass.eduEthicalAI@lemmy.worldEnglish · 1 year agomessage-square0fedilink
EthicalAI@lemmy.worldEnglish · 1 year agoPaLM-E: An embodied multimodal language modelplus-squareai.googleblog.comexternal-linkmessage-square0fedilinkarrow-up14arrow-down12
arrow-up12arrow-down1external-linkPaLM-E: An embodied multimodal language modelplus-squareai.googleblog.comEthicalAI@lemmy.worldEnglish · 1 year agomessage-square0fedilink
ZephyrXero@lemmy.worldEnglish · 1 year agoAlmost All Research on the Mind is in English. That May Be a Problemplus-squarewww.wired.comexternal-linkmessage-square0fedilinkarrow-up16arrow-down12
arrow-up14arrow-down1external-linkAlmost All Research on the Mind is in English. That May Be a Problemplus-squarewww.wired.comZephyrXero@lemmy.worldEnglish · 1 year agomessage-square0fedilink
kromem@lemmy.worldEnglish · 1 year agoLarge language models encode clinical knowledgeplus-squarewww.nature.comexternal-linkmessage-square0fedilinkarrow-up12arrow-down10
arrow-up12arrow-down1external-linkLarge language models encode clinical knowledgeplus-squarewww.nature.comkromem@lemmy.worldEnglish · 1 year agomessage-square0fedilink
DannyMac@lemmy.world · 1 year agoGoogle’s language model “NotebookLM” app hits public testingplus-squarearstechnica.comexternal-linkmessage-square0fedilinkarrow-up15arrow-down10
arrow-up15arrow-down1external-linkGoogle’s language model “NotebookLM” app hits public testingplus-squarearstechnica.comDannyMac@lemmy.world · 1 year agomessage-square0fedilink
DannyMac@lemmy.world · 1 year agoGenerative AI Goes 'MAD' When Trained on AI-Created Data Over Five Timesplus-squarewww.tomshardware.comexternal-linkmessage-square1fedilinkarrow-up17arrow-down11
arrow-up16arrow-down1external-linkGenerative AI Goes 'MAD' When Trained on AI-Created Data Over Five Timesplus-squarewww.tomshardware.comDannyMac@lemmy.world · 1 year agomessage-square1fedilink
DannyMac@lemmy.world · 1 year agoNew ChatGPT rival, Claude 2, launches for open beta testingplus-squarearstechnica.comexternal-linkmessage-square1fedilinkarrow-up16arrow-down10
arrow-up16arrow-down1external-linkNew ChatGPT rival, Claude 2, launches for open beta testingplus-squarearstechnica.comDannyMac@lemmy.world · 1 year agomessage-square1fedilink
kromem@lemmy.worldEnglish · 1 year agoGPT-4 API general availability and deprecation of older models in the Completions APIplus-squareopenai.comexternal-linkmessage-square3fedilinkarrow-up17arrow-down10
arrow-up17arrow-down1external-linkGPT-4 API general availability and deprecation of older models in the Completions APIplus-squareopenai.comkromem@lemmy.worldEnglish · 1 year agomessage-square3fedilink