Uploaded August 2026 | Updated September 2026, 1 week ago
Need high quality cloud GPUs? Check out Verda now, and use the code BYCLOUD-50 to get $50 of compute credits for just $5!
verda.com/?utm_source=bycloud&utm_medium=referral&utm_campaign=sponsorship&utm_content=diffusion_blocks
A great news for us gpu poors? This research DiffusionBlocks shows that you can train LLMs with 2-3x less memory with minimal performance loss. If crazier, you could even go up to 4x or 6x!
my latest project: Intuitive AI Academy
We just wrote a new piece on Optimization!!
https://intuitiveai.academy/
limited time code "LOCKIN" for 35% off yearly plan
My Newsletter (weekly top research papers)
mail.bycloud.ai
My Patreon
patreon.com/c/bycloud
DiffusionBlocks
[Paper] arxiv.org/abs/2506.14202
[Project Page] pub.sakana.ai/diffusionblocks
[Code] github.com/SakanaAI/DiffusionBlocks
Try out my new fav place to learn how to code scrimba.com/?via=bycloudAI
This video is supported by the kind Patrons & YouTube Members:
🙏Spam Maj, Alex, Chris LeDoux, DX Research Group, Poof N' Inu, Deagan, Robert Zawiasa, Ryszard Warzocha, Tobe2d, Louis Muk, Akkusativ, Kevin Tai, Mark Buckler, NO U, Tony Jimenez, Ângelo Fonseca, jiye, Anushka, Asad Dhamani, Binnie Yiu, Calvin Yan, Clayton Ford, Diego Silva, Etrotta, Gonzalo Fidalgo, Handenon, Hector, Jake Disco very, Michael Brenner, Nilly K, OlegWock, Daddy Wen, Shuhong Chen, Sid_Cipher, Stefan Lorenz, Sup, tantan assawade, Thipok Tham, Thomas Di Martino, Thomas Lin, Richárd Nagyfi, Paperboy, mika, Leo, Berhane-Meskel, Kadhai Pesalam, mayssam, Bill Mangrum, nyaa, Toru Mon, Lame Plane, Matej Macak, Len Mo, saylikhapekar, ZyanSheep, THEVIERAOS, Ricardo Raphael Corona-Moreno, superchordate
[Discord] discord.gg/NhJZGtH
[Twitter] twitter.com/bycloudai
[Patreon] patreon.com/bycloud
[Business Inquiries] bycloud@smoothmedia.co
[Other Inquiries] bycloudai@gmail.com
[Profile & Banner Art] twitter.com/pygm7
[Video Editor] @aduckchicken2
Manim Animations created with Manimate manimate.ai
[Ko-fi] ko-fi.com/bycloudai
Need high quality cloud GPUs? Check out Verda now, and use the code BYCLOUD-50 to get $50 of compute credits for just $5!
verda.com/?utm_source=bycloud&utm_medium=referral&utm_campaign=sponsorship&utm_content=diffusion_blocks
A great news for us gpu poors? This research DiffusionBlocks shows that you can train LLMs with 2-3x less memory with minimal performance loss. If crazier, you could even go up to 4x or 6x!
my latest project: Intuitive AI Academy
We just wrote a new piece on Optimization!!
https://intuitiveai.academy/
limited time code "LOCKIN" for 35% off yearly plan
My Newsletter (weekly top research papers)
mail.bycloud.ai
My Patreon
patreon.com/c/bycloud
DiffusionBlocks
[Paper] arxiv.org/abs/2506.14202
[Project Page] pub.sakana.ai/diffusionblocks
[Code] github.com/SakanaAI/DiffusionBlocks
Try out my new fav place to learn how to code scrimba.com/?via=bycloudAI
This video is supported by the kind Patrons & YouTube Members:
🙏Spam Maj, Alex, Chris LeDoux, DX Research Group, Poof N' Inu, Deagan, Robert Zawiasa, Ryszard Warzocha, Tobe2d, Louis Muk, Akkusativ, Kevin Tai, Mark Buckler, NO U, Tony Jimenez, Ângelo Fonseca, jiye, Anushka, Asad Dhamani, Binnie Yiu, Calvin Yan, Clayton Ford, Diego Silva, Etrotta, Gonzalo Fidalgo, Handenon, Hector, Jake Disco very, Michael Brenner, Nilly K, OlegWock, Daddy Wen, Shuhong Chen, Sid_Cipher, Stefan Lorenz, Sup, tantan assawade, Thipok Tham, Thomas Di Martino, Thomas Lin, Richárd Nagyfi, Paperboy, mika, Leo, Berhane-Meskel, Kadhai Pesalam, mayssam, Bill Mangrum, nyaa, Toru Mon, Lame Plane, Matej Macak, Len Mo, saylikhapekar, ZyanSheep, THEVIERAOS, Ricardo Raphael Corona-Moreno, superchordate
[Discord] discord.gg/NhJZGtH
[Twitter] twitter.com/bycloudai
[Patreon] patreon.com/bycloud
[Business Inquiries] bycloud@smoothmedia.co
[Other Inquiries] bycloudai@gmail.com
[Profile & Banner Art] twitter.com/pygm7
[Video Editor] @aduckchicken2
Manim Animations created with Manimate manimate.ai
[Ko-fi] ko-fi.com/bycloudai
![How Googles Transformer 2.0 Might Be The AI Breakthrough We Need
Try Mammouth now for only $10/mo! https://mammouth.ai
In this video, I will be sharing the research that aims to solve the problem of context window, kv-cache, and memory recall efficiency. Even though the title only mentions Googles research, I also included research from Meta and Sakana AI. They paved a good way to introduce to the idea of AI memory.
My Newletter
https://mail.bycloud.ai/
My Patreon
https://www.patreon.com/c/bycloud
An Evolved Universal Transformer Memory
[Paper] https://arxiv.org/abs/2410.13166
Memory Layers at Scale
[Paper] https://arxiv.org/abs/2412.09764
Titans: Learning to Memorize at Test Time
[Paper] https://arxiv.org/abs/2501.00663
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, Ben Shaener, Chris LeDoux, Miguilim, Deagan, FiFaŁ, Robert Zawiasa, Marcelo Ferreira, Owen Ingraham, Daddy Wen, Tony Jimenez, Panther Modern, Jake Disco, Demilson Quintao, Penumbraa, Shuhong Chen, Hongbo Men, happi nyuu nyaa, Carol Lo, Mose Sakashita, Miguel, Bandera, Gennaro Schiano, gunwoo, Ravid Freedman, Mert Seftali, Mrityunjay, Richárd Nagyfi, Timo Steiner, Henrik G Sundt, projectAnthony, Brigham Hall, Kyle Hudson, Kalila, Jef Come, Jvari Williams, Tien Tien, BIll Mangrum, owned, Janne Kytölä, SO, Richárd Nagyfi, Hector, Drexon, Claxvii 177th, Inferencer, Michael Brenner, Akkusativ, Oleg Wock, FantomBloth, Thipok Tham, Clayton Ford, Theo, Handenon, Diego Silva, mayssam, Kadhai Pesalam, Tim Schulz, jiye, Anushka, Henrik Sundt, Julian Aßmann, Thomas Lin, Sid_Cypher, Mark Buckler, Kevin Tai, NO U, Gonzalo Fidalgo, Igor Alvarez, Alon Pluda, Clément Veyssière, Sander Zwaenepoel, etrotta, Binnie Yiu, Matej Macak, c zhou, Berhane-Meskel, sai sandeep mandava, Leo, Asad Dhamani, Charlie C, tantan assawade, Ângelo Fonseca, Stefan Lorenz, Paperboy, mika, Leo, Utsav Soi
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Business Inquiries] bycloudai@gmail.com
[Music] massobeats - lavender
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] @Booga04
[3D Animation] @F_E_U
[Bitcoin (BTC)] 3JFMJQVGXNA2HJE5V9qCwLiqy6wHY9Vhdx
[Ethereum (ETH)] 0x3d784F55E0bE5f35c1566B2E014598C0f354f190
[Litecoin (LTC)] MGHnqALjyU2W6NuJSSW9fTWV4dcHfwHZd7
[Bitcoin Cash (BCH)] 1LkyGfzHxnSfqMF8tN7ZGDwUTyBB6vcii9
[Solana (SOL)] 6XyMCEdVhtxJQRjMKgUJaySL8cGoBPzzA2NPDMPfVkKN
[Ko-fi] https://ko-fi.com/bycloudai How Googles Transformer 2.0 Might Be The AI Breakthrough We Need](https://i.ytimg.com/vi/UMkCmOTX5Ow/mqdefault.jpg)

![The Chinese AI Iceberg
Dont miss out on this FREE HubSpot for Startups Exclusive AI Insights! https://clickhubspot.com/682e44
My Newsletter
https://mail.bycloud.ai/
my project: find, discover & explain AI research semantically
https://findmypapers.ai/
My Patreon
https://www.patreon.com/c/bycloud
In this video, I will break down the current Chinese AI landscape from flagship labs to underground research groups: DeepSeek, Alibaba Cloud’s Qwen, ByteDance Seed, Tencents Hunyuan, Baidu’s ERNI, Moonshot AI’s Kimi K2, Zhipu AI (Z AI) and its GLM series, MiniMax, Kuaishou with KLING, Huawei’s Pangus, SenseTime’s SenseNova, Shanghai AI Laboratory with InternLM and InternViT, Ant Group’s Ring and Inclusion AI, XiaoHongShu’s dots.llm1, Xiaomi’s MiMO family, and Meituan LongCat.
Links
[Artificial Analysis] https://artificialanalysis.ai/
[DeepSeek] https://github.com/deepseek-ai/DeepSeek-V3
[Qwen] https://chat.qwen.ai/
[ByteDance Seed] https://seed.bytedance.com/en/
[Tencent Hunyuan] https://hunyuan.tencent.com/
[Kimi] https://www.kimi.com/
[MiniMax M2] https://www.minimaxi.com/news/minimax-m2
[Z AI] https://chat.z.ai/
[Manus] https://manus.im/app
[KLING] https://klingai.com/global/
[ERINE] https://ernie.baidu.com/
[Baichuan] https://ying.baichuan-ai.com/chat
[01 AI] https://www.01.ai/
[OpenBMB] https://www.openbmb.cn/en
[Noahs Ark Lab] https://noahs-ark.github.io/
[Noahs Ark Lab Controversy] https://github.com/HW-whistleblower/True-Story-of-Pangu
[SenseTime] https://www.sensetime.com/hk
[Shanghais AI Lab] https://www.shlab.org.cn/
[Ant Group] https://www.antgroup.com/en
[Cherry Studio] https://www.cherry-ai.com/
[Rednote] https://huggingface.co/rednote-hilab
[Xiaomi] https://huggingface.co/XiaomiMiMo
[LongCat] https://huggingface.co/meituan-longcat
I am not affiliated nor sponsored by any of these AI labs/companies I mentioned in this video. This videos sponsor is solely HubSpot for Startups.
Try out my new fav place to learn how to code https://scrimba.com/?via=bycloudAI
This video is supported by the kind Patrons & YouTube Members:
🙏Spam Maj, Alex, Chris LeDoux, DX Research Group, Poof N Inu, Deagan, Robert Zawiasa, Ryszard Warzocha, Tobe2d, Louis Muk, Akkusativ, Kevin Tai, Mark Buckler, NO U, Tony Jimenez, Ângelo Fonseca, jiye, Anushka, Asad Dhamani, Binnie Yiu, Calvin Yan, Clayton Ford, Diego Silva, Etrotta, Gonzalo Fidalgo, Handenon, Hector, Jake Disco very, Michael Brenner, Nilly K, OlegWock, Daddy Wen, Shuhong Chen, Sid_Cipher, Stefan Lorenz, Sup, tantan assawade, Thipok Tham, Thomas Di Martino, Thomas Lin, Richárd Nagyfi, Paperboy, mika, Leo, Berhane-Meskel, Kadhai Pesalam, mayssam, Bill Mangrum, nyaa, Toru Mon
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Business Inquiries] bycloud@smoothmedia.co
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] @Booga04
[Ko-fi] https://ko-fi.com/bycloudai
0:00 Intro
1:53 DeepSeek
3:21 Qwen
4:57 ByteDance Seed
6:48 Hunyuan
8:06 Kimi
9:16 Zhipu AI (Z.AI)
10:19 MiniMax
12:11 Kuaishou
12:43 Manus AI
14:05 Baidu ERINE
15:15 01.AI
16:11 Baichuan
16:35 StepFun
17:33 OpenBMB
18:22 Huaweis Noahs Ark Lab
19:53 SenseTime
20:50 Shanghai AI Labs
21:56 Ant Group
23:05 Cherry Studio
23:42 Rednote
24:18 Xiaomi
24:55 Meituan LongCat The Chinese AI Iceberg](https://i.ytimg.com/vi/XFhUI1fphKU/mqdefault.jpg)
![Claude 3.7 Sonnet: The AI Code King Has Returned (within 1 week)
My Newletter
https://mail.bycloud.ai/
My Patreon
https://www.patreon.com/c/bycloud
Here is the first look of Claude 3.7 Sonnet. My copium is high, it is looking good. I will return to the Claude 3.7 Sonnet 20$ subscription.
Claude 3.7 Sonnet
[Full Blog] https://www.anthropic.com/news/claude-3-7-sonnet
[System Card] https://assets.anthropic.com/m/785e231869ea8b3b/original/claude-3-7-sonnet-system-card.pdf
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, Ben Shaener, Chris LeDoux, Miguilim, Deagan, FiFaŁ, Robert Zawiasa, Marcelo Ferreira, Owen Ingraham, Daddy Wen, Tony Jimenez, Panther Modern, Jake Disco, Demilson Quintao, Penumbraa, Shuhong Chen, Hongbo Men, happi nyuu nyaa, Carol Lo, Mose Sakashita, Miguel, Bandera, Gennaro Schiano, gunwoo, Ravid Freedman, Mert Seftali, Mrityunjay, Richárd Nagyfi, Timo Steiner, Henrik G Sundt, projectAnthony, Brigham Hall, Kyle Hudson, Kalila, Jef Come, Jvari Williams, Tien Tien, BIll Mangrum, owned, Janne Kytölä, SO, Richárd Nagyfi, Hector, Drexon, Claxvii 177th, Inferencer, Michael Brenner, Akkusativ, Oleg Wock, FantomBloth, Thipok Tham, Clayton Ford, Theo, Handenon, Diego Silva, mayssam, Kadhai Pesalam, Tim Schulz, jiye, Anushka, Henrik Sundt, Julian Aßmann, Thomas Lin, Sid_Cypher, Mark Buckler, Kevin Tai, NO U, Gonzalo Fidalgo, Igor Alvarez, Alon Pluda, Clément Veyssière, Sander Zwaenepoel, etrotta, Binnie Yiu, Matej Macak, c zhou, Berhane-Meskel, sai sandeep mandava, Leo, Asad Dhamani, Charlie C, tantan assawade, Ângelo Fonseca, Stefan Lorenz, Paperboy, mika, Leo, Utsav Soi
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Business Inquiries] bycloudai@gmail.com
[Music]
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] @Booga04 and me
[Bitcoin (BTC)] 3JFMJQVGXNA2HJE5V9qCwLiqy6wHY9Vhdx
[Ethereum (ETH)] 0x3d784F55E0bE5f35c1566B2E014598C0f354f190
[Litecoin (LTC)] MGHnqALjyU2W6NuJSSW9fTWV4dcHfwHZd7
[Bitcoin Cash (BCH)] 1LkyGfzHxnSfqMF8tN7ZGDwUTyBB6vcii9
[Solana (SOL)] 6XyMCEdVhtxJQRjMKgUJaySL8cGoBPzzA2NPDMPfVkKN
[Ko-fi] https://ko-fi.com/bycloudai Claude 3.7 Sonnet: The AI Code King Has Returned (within 1 week)](https://i.ytimg.com/vi/XtD0pG6h_gM/mqdefault.jpg)
![The LK-99 of AI: The Reflection-70B Controversy Full Rundown
For people wondering why I draw similarity with LK-99, its because the results are not reproducible.
The saga of reflection-70B has been a wild one. This video, I have made a full break down on the situation, with the latest up-to-date information, and why we probably wont be expecting anything good out from it anymore.
check out my newsletter:
https://mail.bycloud.ai/
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, alex j, Chris LeDoux, Alex Maurice, Miguilim, Deagan, FiFaŁ, Robert Zawiasa, Owen Ingraham, Daddy Wen, Tony Jimenez, Panther Modern, Jake Disco, Demilson Quintao, Penumbraa, Shuhong Chen, Hongbo Men, happi nyuu nyaa, Carol Lo, Mose Sakashita, Miguel, Bandera, Gennaro Schiano, gunwoo, Ravid Freedman, Mert Seftali, Mrityunjay, Richárd Nagyfi, Timo Steiner, Henrik G Sundt, projectAnthony, Brigham Hall, Kyle Hudson, Kalila, Jef Come, Jvari Williams, Tien Tien, BIll Mangrum, owned, Janne Kytölä, SO, Richárd Nagyfi, Hector, Drexon, Claxvii 177th, Inferencer, Michael Brenner, Akkusativ, Oleg Wock, FantomBloth, Thipok Tham, Clayton Ford, Theo, Handenon
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Music] Sunny Days - Anno Domini Beats
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] @HiMyNameisKieran The LK-99 of AI: The Reflection-70B Controversy Full Rundown](https://i.ytimg.com/vi/Xtr_Ll_A9ms/mqdefault.jpg)
![All You Need To Know About Running LLMs Locally
my latest project: Intuitive AI Academy, learn modern AI/LLMs Intuitively
https://intuitiveai.academy/
code NYNM for 50% off forever (limited to 50)
TensorRT LLM
[Code] https://github.com/NVIDIA/TensorRT-LLM
[Getting Started Blog] https://nvda.ws/3O7f8up
[Dev Blog] https://nvda.ws/490uadi
Chat with RTX
[Download] https://nvda.ws/3OHPRHE
[Blog] https://nvda.ws/3whKZTb
Links:
[Oobabooga] https://github.com/oobabooga/text-generation-webui
[SillyTavern] https://github.com/SillyTavern/SillyTavern
[LM Studio] https://lmstudio.ai/
[Axolotl] https://github.com/OpenAccess-AI-Collective/axolotl
[Llama Factory] https://github.com/hiyouga/LLaMA-Factory
[HuggingFace] https://huggingface.co/models
[AWQ] https://github.com/mit-han-lab/llm-awq
[ExLlamav2] https://github.com/turboderp/exllamav2
[GGUF] https://github.com/ggerganov/ggml/blob/master/docs/gguf.md
[GPTQ] https://github.com/IST-DASLab/gptq
[LlamaCpp] https://github.com/ggerganov/llama.cpp
[vllm] https://github.com/vllm-project/vllm
[TensorRT LLM] https://github.com/NVIDIA/TensorRT-LLM
[Chat with RTX] https://www.nvidia.com/en-us/ai-on-rtx/chat-with-rtx-generative-ai/
[LlamaIndex] https://github.com/run-llama/llama_index
[Continue.dev] https://continue.dev/
Model recommendations (I know you are here after DeepSeek):
[All DeepSeek Models] https://huggingface.co/collections/deepseek-ai/deepseek-r1-678e1e131c0169c0bc89728d
[Easily Download with Ollama] https://ollama.com/library/deepseek-r1
Heres the rule of thumb to know if you can run it:
If your VRAM is larger than the model GB size * 1.2, than you can run that model size locally.
Eg. DeepSeek-7B = 4.7GB then 4.7*1.2=5.64, so if your GPU has 8GB VRAM, since 8GB is bigger than 5.64, you can run DeepSeek-7B.
Check out my latest video on DeepSeek-R1 to understand the context better!
(the following are all outdated)
Just use Llama-3.1 instead for everything.
[Llama-3.1] https://huggingface.co/collections/meta-llama/llama-31-669fc079a0c406a149a5738f
Translation can try Aya 23
[Aya 23] https://huggingface.co/CohereForAI/aya-23-8B
(the following are all outdated)
[Nous-Hermes-llama-2-7b] https://huggingface.co/NousResearch/Nous-Hermes-llama-2-7b
[Openchat-3.5-0106] https://huggingface.co/openchat/openchat-3.5-0106
[SOLAR-10.7B-Instruct-v1.0] https://huggingface.co/upstage/SOLAR-10.7B-Instruct-v1.0
[Google Gemma] https://huggingface.co/google/gemma-7b
[Mixtral-8x7B-Instruct-v0.1] https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1
[Deepseek-coder-33b-instruct] https://huggingface.co/deepseek-ai/deepseek-coder-33b-instruct
[Colbertv2.0] https://huggingface.co/colbert-ir/colbertv2.0
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, alex j, Chris LeDoux, Alex Maurice, Miguilim, Deagan, FiFaŁ, Daddy Wen, Tony Jimenez, Panther Modern, Jake Disco, Demilson Quintao, Shuhong Chen, Hongbo Men, happi nyuu nyaa, Carol Lo, Mose Sakashita, Miguel, Bandera, Gennaro Schiano, gunwoo, Ravid Freedman, Mert Seftali, Mrityunjay, Richárd Nagyfi, Timo Steiner, Henrik G Sundt, projectAnthony, Brigham Hall, Kyle Hudson, Kalila, Jef Come, Jvari Williams, Tien Tien, BIll Mangrum, owned, Janne Kytölä, SO, Richárd Nagyfi, Hector, Drexon
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Music] massobeats - magic carousel
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] maikadihaika All You Need To Know About Running LLMs Locally](https://i.ytimg.com/vi/XwL_cRuXM2E/mqdefault.jpg)
![The Difference of AI Videos No One Tells You About
check out DomoAI and start your vid2vid experiments! https://r.domoai.app/bycloud
Every type of AI generated video ever...? Well this video is definitely a bit more ambitious but I think I covered most of the popular AI processed/generated videos there possibly are. Now I think about it, most of it that went mainstream are because of memes lmao
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, alex j, Chris LeDoux, Alex Maurice, Miguilim, Deagan, FiFaŁ, Robert Zawiasa, Owen Ingraham, Daddy Wen, Tony Jimenez, Panther Modern, Jake Disco, Demilson Quintao, Penumbraa, Shuhong Chen, Hongbo Men, happi nyuu nyaa, Carol Lo, Mose Sakashita, Miguel, Bandera, Gennaro Schiano, gunwoo, Ravid Freedman, Mert Seftali, Mrityunjay, Richárd Nagyfi, Timo Steiner, Henrik G Sundt, projectAnthony, Brigham Hall, Kyle Hudson, Kalila, Jef Come, Jvari Williams, Tien Tien, BIll Mangrum, owned, Janne Kytölä, SO, Richárd Nagyfi, Hector, Drexon, Claxvii 177th, Inferencer, Michael Brenner, Akkusativ, Oleg Wock, FantomBloth, Thipok Tham, Clayton Ford, Theo, Handenon
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Music] massobeats - aromatic
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] @Askejm The Difference of AI Videos No One Tells You About](https://i.ytimg.com/vi/Y43EgHv0xrs/mqdefault.jpg)
![Attention Sink: The Fluke That Made LLMs Actually Usable
Get started now with privacy focused VPN by Proton! https://proton.me/pass/bycloudai
My Newletter
https://mail.bycloud.ai/
My Patreon
https://www.patreon.com/c/bycloud
Efficient Streaming Language Models with Attention Sinks
[Paper] https://arxiv.org/abs/2309.17453
Why do LLMs attend to the first token?
[Paper] https://arxiv.org/abs/2504.02732
Softmax Attention is a Fluke
[Blog] https://www.ethansmith2000.com/post/softmax-attention-is-a-fluke
If you want to learn more about Attention Sink, check out my latest project
https://findmypapers.ai/
Where you can ask, find & have it explain AI research that also discuss about Attention Sink!
Try out my new fav place to learn how to code https://scrimba.com/?via=bycloudAI
This video is supported by the kind Patrons & YouTube Members:
🙏Chris LeDoux, DX Research Group, Poof N Inu, Deagan, Robert Zawiasa, Tobe2d, Louis Muk, Akkusativ, Kevin Tai, Mark Buckler, NO U, Tony Jimenez, Ângelo Fonseca, jiye, Anushka, Asad Dhamani, Binnie Yiu, Calvin Yan, Clayton Ford, Diego Silva, Etrotta, Gonzalo Fidalgo, Handenon, Hector, Jake Disco very, Michael Brenner, Nilly K, OlegWock, Daddy Wen, Shuhong Chen, Sid_Cipher, Stefan Lorenz, Sup, tantan assawade, Thipok Tham, Thomas Di Martino, Thomas Lin, Richárd Nagyfi, Paperboy, mika, Leo, Berhane-Meskel, Kadhai Pesalam, mayssam, Bill Mangrum, nyaa
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Business Inquiries] bycloud@smoothmedia.co
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] Abhay
[Ko-fi] https://ko-fi.com/bycloudai Attention Sink: The Fluke That Made LLMs Actually Usable](https://i.ytimg.com/vi/Y8Tj9kq4iWY/mqdefault.jpg)
![The AI Hardware Arms Race Is Getting Out of Hand
Check out Gamma.app now using this link here!
https://gamma.1stcollab.com/bycloud
CS-3
[Blog] https://cerebras.net/press-release/cerebras-announces-third-generation-wafer-scale-engine
Groq
[Groq online Inference] https://groq.com/
[Groq LPU Blog] https://groq.com/wp-content/uploads/2022/10/GroqChip%E2%84%A2-Processor-Product-Brief-v1.5.pdf
Truffle-1
[Store] https://preorder.itsalltruffles.com/
Intel Max1550 Paper
[paper] https://arxiv.org/abs/2403.17607
[Tweet] https://x.com/main_horse/status/1772816958167081123
Extropic
[Blog] https://www.extropic.ai/future
[Andrews Explainer] https://x.com/Andercot/status/1767252816660471878
[Documentary] https://x.com/jasonjoyride/status/1784291509246001197
[Documentary Retweet] https://x.com/0xKyon/status/1784591427462389822
[Tweet Drama Thread 1] https://x.com/BasedBeffJezos/status/1768349738271936591
[Tweet Drama Thread 2] https://x.com/fluxtheorist/status/1784645914633662844
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, alex j, Chris LeDoux, Alex Maurice, Miguilim, Deagan, FiFaŁ, Daddy Wen, Tony Jimenez, Panther Modern, Jake Disco, Demilson Quintao, Shuhong Chen, Hongbo Men, happi nyuu nyaa, Carol Lo, Mose Sakashita, Miguel, Bandera, Gennaro Schiano, gunwoo, Ravid Freedman, Mert Seftali, Mrityunjay, Richárd Nagyfi, Timo Steiner, Henrik G Sundt, projectAnthony, Brigham Hall, Kyle Hudson, Kalila, Jef Come, Jvari Williams, Tien Tien, BIll Mangrum, owned, Janne Kytölä, SO, Richárd Nagyfi, Hector, Drexon, Claxvii 177th, Inferencer, Michael Brenner, Akkusativ, Oleg Wock
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Music] Massobeats - Glisten
[Music] Massobeats - Lush
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] Silas
0:00 Intro
0:35 CS-3
1:52 Groq
2:48 Truffle-1
4:09 Intel Max 1550 Research
5:18 Extropic Chip
8:08 Gamma.app (sponsor) The AI Hardware Arms Race Is Getting Out of Hand](https://i.ytimg.com/vi/YAGdtvHGZaA/mqdefault.jpg)

![Has ChatGPT Finally Been Dethroned? Claude 3 Review
Want to compare the results from the chatbots yourself? Check out ChatHub now! They offer one of the easiest ways for you to compare and evaluate the results to find the best LLM for you:
https://chathub.gg/?utm_source=bycloud
In this video, Ill be sharing with you my personal experiences from using Claude 3 for probably non-stop 3 weeks. I feel like I might have glazed claude 3 a bit too hard, but I actually have been using it for coding most of the time.
Claude 3
[Webpage] https://claude.ai/
[Blog] https://www.anthropic.com/news/claude-3-family
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, alex j, Chris LeDoux, Alex Maurice, Miguilim, Deagan, FiFaŁ, Daddy Wen, Tony Jimenez, Panther Modern, Jake Disco, Demilson Quintao, Shuhong Chen, Hongbo Men, happi nyuu nyaa, Carol Lo, Mose Sakashita, Miguel, Bandera, Gennaro Schiano, gunwoo, Ravid Freedman, Mert Seftali, Mrityunjay, Richárd Nagyfi, Timo Steiner, Henrik G Sundt, projectAnthony, Brigham Hall, Kyle Hudson, Kalila, Jef Come, Jvari Williams, Tien Tien, BIll Mangrum, owned, Janne Kytölä, SO, Richárd Nagyfi, Hector, Drexon, Claxvii 177th, Inferencer
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] @askejm Has ChatGPT Finally Been Dethroned? Claude 3 Review](https://i.ytimg.com/vi/YfMHmhWWHV0/mqdefault.jpg)