☆ Yσɠƚԋσʂ ☆English · 2 个月前The Attention Mechanism Born for Cost Optimizationplus-squareoilbeater.comexternal-linkmessage-square0linkfedilinkarrow-up16arrow-down10
arrow-up16arrow-down1external-linkThe Attention Mechanism Born for Cost Optimizationplus-squareoilbeater.com☆ Yσɠƚԋσʂ ☆English · 2 个月前message-square0linkfedilink
thickertoofan@lemm.eeEnglish · 2 个月前dcdaML - devanagari character detection dataset training frameworkplus-squaregithub.comexternal-linkmessage-square5linkfedilinkarrow-up15arrow-down10
arrow-up15arrow-down1external-linkdcdaML - devanagari character detection dataset training frameworkplus-squaregithub.comthickertoofan@lemm.eeEnglish · 2 个月前message-square5linkfedilink
☆ Yσɠƚԋσʂ ☆English · 3 个月前Neural Graffiti is an experiment in adding a "Spray Layer" to a transformer model, which injects a memory trace into the final stages of inference without finetuning or retrainingplus-squaregithub.comexternal-linkmessage-square0linkfedilinkarrow-up15arrow-down10
arrow-up15arrow-down1external-linkNeural Graffiti is an experiment in adding a "Spray Layer" to a transformer model, which injects a memory trace into the final stages of inference without finetuning or retrainingplus-squaregithub.com☆ Yσɠƚԋσʂ ☆English · 3 个月前message-square0linkfedilink
fubarx@lemmy.world · 3 个月前Breaking GPT-5 News!plus-squaremessage-squaremessage-square2linkfedilinkarrow-up15arrow-down111
arrow-up1-6arrow-down1message-squareBreaking GPT-5 News!plus-squarefubarx@lemmy.world · 3 个月前message-square2linkfedilink
4Robato@lemmy.worldEnglish · edit-23 个月前I want to open source a dataset but I'm not sure what license to useplus-squaremessage-squaremessage-square6linkfedilinkarrow-up14arrow-down10
arrow-up14arrow-down1message-squareI want to open source a dataset but I'm not sure what license to useplus-square4Robato@lemmy.worldEnglish · edit-23 个月前message-square6linkfedilink
☆ Yσɠƚԋσʂ ☆English · 3 个月前Why do LLMs make stuff up? New research peers under the hood.plus-squarearstechnica.comexternal-linkmessage-square0linkfedilinkarrow-up17arrow-down11
arrow-up16arrow-down1external-linkWhy do LLMs make stuff up? New research peers under the hood.plus-squarearstechnica.com☆ Yσɠƚԋσʂ ☆English · 3 个月前message-square0linkfedilink
oba@lemmy.worldEnglish · 3 个月前MLOps tips I gathered recentlyplus-squarewww.readyforagents.comexternal-linkmessage-square0linkfedilinkarrow-up18arrow-down10
arrow-up18arrow-down1external-linkMLOps tips I gathered recentlyplus-squarewww.readyforagents.comoba@lemmy.worldEnglish · 3 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 4 个月前DeepSeek open source DeepEP – library for MoE training and Inferenceplus-squaregithub.comexternal-linkmessage-square0linkfedilinkarrow-up111arrow-down10
arrow-up111arrow-down1external-linkDeepSeek open source DeepEP – library for MoE training and Inferenceplus-squaregithub.com☆ Yσɠƚԋσʂ ☆English · 4 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 4 个月前Towards Monosemanticity: Decomposing Language Models With Dictionary Learningtransformer-circuits.pubexternal-linkmessage-square0linkfedilinkarrow-up15arrow-down11
arrow-up14arrow-down1external-linkTowards Monosemanticity: Decomposing Language Models With Dictionary Learningtransformer-circuits.pub☆ Yσɠƚԋσʂ ☆English · 4 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 4 个月前Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnetplus-squaretransformer-circuits.pubexternal-linkmessage-square0linkfedilinkarrow-up14arrow-down10
arrow-up14arrow-down1external-linkScaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnetplus-squaretransformer-circuits.pub☆ Yσɠƚԋσʂ ☆English · 4 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 5 个月前DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningarxiv.orgexternal-linkmessage-square0linkfedilinkarrow-up17arrow-down10
arrow-up17arrow-down1external-linkDeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningarxiv.org☆ Yσɠƚԋσʂ ☆English · 5 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 6 个月前Neurosymbolic AI -- Why, What, and Howplus-squarearxiv.orgexternal-linkmessage-square0linkfedilinkarrow-up18arrow-down10
arrow-up18arrow-down1external-linkNeurosymbolic AI -- Why, What, and Howplus-squarearxiv.org☆ Yσɠƚԋσʂ ☆English · 6 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 6 个月前Classical Sorting Algorithms as a Model of Morphogenesis: self-sorting arrays reveal unexpected competencies in a minimal model of basal intelligenceplus-squarearxiv.orgexternal-linkmessage-square0linkfedilinkarrow-up13arrow-down10
arrow-up13arrow-down1external-linkClassical Sorting Algorithms as a Model of Morphogenesis: self-sorting arrays reveal unexpected competencies in a minimal model of basal intelligenceplus-squarearxiv.org☆ Yσɠƚԋσʂ ☆English · 6 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 7 个月前Genie 2: A large-scale foundation world modelplus-squaredeepmind.googleexternal-linkmessage-square1linkfedilinkarrow-up16arrow-down10
arrow-up16arrow-down1external-linkGenie 2: A large-scale foundation world modelplus-squaredeepmind.google☆ Yσɠƚԋσʂ ☆English · 7 个月前message-square1linkfedilink
☆ Yσɠƚԋσʂ ☆English · 7 个月前A good primer on what to expect running local LLMsplus-squarenullprogram.comexternal-linkmessage-square4linkfedilinkarrow-up15arrow-down10
arrow-up15arrow-down1external-linkA good primer on what to expect running local LLMsplus-squarenullprogram.com☆ Yσɠƚԋσʂ ☆English · 7 个月前message-square4linkfedilink
Shamar@feddit.itEnglish · edit-28 个月前A community statement supporting the Open Source Definition (OSD)plus-squareosd.fyiexternal-linkmessage-square1linkfedilinkarrow-up111arrow-down11
arrow-up110arrow-down1external-linkA community statement supporting the Open Source Definition (OSD)plus-squareosd.fyiShamar@feddit.itEnglish · edit-28 个月前message-square1linkfedilink
☆ Yσɠƚԋσʂ ☆English · 9 个月前How ‘Embeddings’ Encode What Words Meanplus-squarewww.quantamagazine.orgexternal-linkmessage-square0linkfedilinkarrow-up19arrow-down10
arrow-up19arrow-down1external-linkHow ‘Embeddings’ Encode What Words Meanplus-squarewww.quantamagazine.org☆ Yσɠƚԋσʂ ☆English · 9 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 10 个月前New AI model “learns” how to simulate Super Mario Bros. from video footageplus-squarearstechnica.comexternal-linkmessage-square0linkfedilinkarrow-up14arrow-down10
arrow-up14arrow-down1external-linkNew AI model “learns” how to simulate Super Mario Bros. from video footageplus-squarearstechnica.com☆ Yσɠƚԋσʂ ☆English · 10 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 10 个月前Reflection 70B holds its own against even the top closed-source models (Claude 3.5 Sonnet, GPT-4o)plus-squarehuggingface.coexternal-linkmessage-square0linkfedilinkarrow-up115arrow-down10
arrow-up115arrow-down1external-linkReflection 70B holds its own against even the top closed-source models (Claude 3.5 Sonnet, GPT-4o)plus-squarehuggingface.co☆ Yσɠƚԋσʂ ☆English · 10 个月前message-square0linkfedilink
☆ Yσɠƚԋσʂ ☆English · 10 个月前It’s Not Intelligent If It Always Halts: A Critical Perspective on Current Approaches to AGIplus-squarewww.lifeiscomputation.comexternal-linkmessage-square6linkfedilinkarrow-up110arrow-down10
arrow-up110arrow-down1external-linkIt’s Not Intelligent If It Always Halts: A Critical Perspective on Current Approaches to AGIplus-squarewww.lifeiscomputation.com☆ Yσɠƚԋσʂ ☆English · 10 个月前message-square6linkfedilink