Uploaded March 2025 | Updated September 2026, 1 week ago
Remove your personal information from the web at JoinDeleteMe.com/BYCLOUD and use
code BYCLOUD for 20% off🙌
DeleteMe international Plans: international.joindeleteme.com
In this video, I will be going over the current problems of chain-of-thoughts for the token based reasoning LLMs, and share with you the next forefront of reasoning that doesn't use any extra tokens or words anymore.
My Newletter
mail.bycloud.ai
My Patreon
patreon.com/c/bycloud
CoCoNuT: Training Large Language Models to Reason in a Continuous Latent Space
[Paper] arxiv.org/abs/2412.06769
Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
[Paper] arxiv.org/abs/2502.05171
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, Ben Shaener, Kainan, 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] discord.gg/NhJZGtH
[Twitter] twitter.com/bycloudai
[Patreon] patreon.com/bycloud
[Business Inquiries] bycloudai@gmail.com
[Profile & Banner Art] twitter.com/pygm7
[Video Editor] Abhay
[Bitcoin (BTC)] 3JFMJQVGXNA2HJE5V9qCwLiqy6wHY9Vhdx
[Ethereum (ETH)] 0x3d784F55E0bE5f35c1566B2E014598C0f354f190
[Litecoin (LTC)] MGHnqALjyU2W6NuJSSW9fTWV4dcHfwHZd7
[Bitcoin Cash (BCH)] 1LkyGfzHxnSfqMF8tN7ZGDwUTyBB6vcii9
[Solana (SOL)] 6XyMCEdVhtxJQRjMKgUJaySL8cGoBPzzA2NPDMPfVkKN
[Ko-fi] ko-fi.com/bycloudai
Remove your personal information from the web at JoinDeleteMe.com/BYCLOUD and use
code BYCLOUD for 20% off🙌
DeleteMe international Plans: international.joindeleteme.com
In this video, I will be going over the current problems of chain-of-thoughts for the token based reasoning LLMs, and share with you the next forefront of reasoning that doesn't use any extra tokens or words anymore.
My Newletter
mail.bycloud.ai
My Patreon
patreon.com/c/bycloud
CoCoNuT: Training Large Language Models to Reason in a Continuous Latent Space
[Paper] arxiv.org/abs/2412.06769
Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
[Paper] arxiv.org/abs/2502.05171
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, Ben Shaener, Kainan, 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] discord.gg/NhJZGtH
[Twitter] twitter.com/bycloudai
[Patreon] patreon.com/bycloud
[Business Inquiries] bycloudai@gmail.com
[Profile & Banner Art] twitter.com/pygm7
[Video Editor] Abhay
[Bitcoin (BTC)] 3JFMJQVGXNA2HJE5V9qCwLiqy6wHY9Vhdx
[Ethereum (ETH)] 0x3d784F55E0bE5f35c1566B2E014598C0f354f190
[Litecoin (LTC)] MGHnqALjyU2W6NuJSSW9fTWV4dcHfwHZd7
[Bitcoin Cash (BCH)] 1LkyGfzHxnSfqMF8tN7ZGDwUTyBB6vcii9
[Solana (SOL)] 6XyMCEdVhtxJQRjMKgUJaySL8cGoBPzzA2NPDMPfVkKN
[Ko-fi] ko-fi.com/bycloudai
![So Gaslighting AI Is A Thing Now [ChatGPT]](https://i.ytimg.com/vi/rKgb5t0_vsM/mqdefault.jpg "So Gaslighting AI Is A Thing Now [ChatGPT]
Movio is a web app that lets you generate an AI spokesperson with whatever text you write and lip sync it for you perfectly. Try it now for free with the FREE avatar lite package that youll exclusively get with the link here (ad):
https://app.movio.la/guest?sid=bycloud
Do I even need a chatGPT link here anymore? Everyone that knows it should have used it by now lol
https://chat.openai.com/chat
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, Chris LeDoux, Panther Modern, Jake Disco, Demilson Quintao, Tony Jimenez, 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, tianmuhui, projectAnthony
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[PayPal] https://paypal.me/bycloudai
[Profile & Banner Art] https://twitter.com/pygm7
[Video Editor] askejm{hashtag}1111 on Discord
#chatgpt #movio #talkingavatar #custom avatar")
![Apple Joins The Extended Reality Development [The AI Timeline #4]](https://i.ytimg.com/vi/s64DwZlvR2U/mqdefault.jpg "Apple Joins The Extended Reality Development [The AI Timeline #4]
Welcome to my 4th episode of the weekly AI news series The AI Timeline, where I go through the AI news in the past week with the most distilled information and memes!
QR Code ControlNET (mb I dont think its a model, its just ref image)
[Reddit] https://www.reddit.com/r/StableDiffusion/comments/141hg9x/controlnet_for_qr_code/
Safe.ai Statement
[Official Website] https://www.safe.ai/statement-on-ai-risk
Andrew Ng Statement
[TW] https://twitter.com/AndrewYNg/status/1665759430552567810?s=20
SAM-HQ
[Paper] https://arxiv.org/abs/2306.01567
[Code] https://github.com/SysCV/SAM-HQ
SAM
[Project Page] https://segment-anything.com/
Potat1 (Open Source Text 2 Video Model)
[HuggingFace] https://huggingface.co/camenduru/potat1
“Intelligent brains take longer to solve difficult problems”
[Blog] https://www.bihealth.org/en/notices/intelligent-brains-take-longer-to-solve-difficult-problems
[Paper] https://www.nature.com/articles/s41467-023-38626-y
StyleDrop
[Project Page] https://styledrop.github.io/
AI Chatbot Paper on these attention-seeking LLMs
[Paper] https://arxiv.org/abs/2303.06135
Japan Government on AI (or AI Art) Training
[TW] https://twitter.com/noe_emk/status/1664613177106239488?s=20
[Article] https://technomancers.ai/japan-goes-all-in-copyright-doesnt-apply-to-ai-training/#more-642
[JP Article] https://pc.watch.impress.co.jp/docs/news/1506018.html
Israel Government on AI (or AI Art) Training
[Article] https://www.project-disco.org/intellectual-property/011823-israel-ministry-of-justice-issues-opinion-supporting-the-use-of-copyrighted-works-for-machine-learning/
Time Magazine New Cover
[TW] https://twitter.com/TIME/status/1663939590908985348?s=20
OpenAIs 1mil Grant For Cybersecurity
[Blog] https://openai.com/blog/openai-cybersecurity-grant-program
Snap Fusion
[HuggingFace] https://huggingface.co/papers/2306.00980
[Paper] https://arxiv.org/abs/2306.00980
[Project Page] https://snap-research.github.io/SnapFusion/
Bytes Are All You Need
[Paper] https://arxiv.org/abs/2306.00238
[Huggface] https://huggingface.co/papers/2306.00238
OpenAIs Improved Mathematical Reasoning
[Blog] https://openai.com/research/improving-mathematical-reasoning-with-process-supervision
Redream
[GitHub] https://github.com/Fictiverse/Redream
GPTFille (GPT File Organization)
[GitHub] https://github.com/aaronn/gptfile
[TW] https://twitter.com/localghost/status/1663274587860393984?s=20
Runway Gen2
[TW] https://twitter.com/runwayml/status/1666429706932043776?s=20
This video is supported by the kind Patrons & YouTube Members:
🙏Andrew Lescelius, Chris LeDoux, Alex Maurice, 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
[Discord] https://discord.gg/NhJZGtH
[Twitter] https://twitter.com/bycloudai
[Patreon] https://www.patreon.com/bycloud
[Profile & Banner Art] https://twitter.com/pygm7
Chapters:
0:00 Apples VisionPro
1:07 AI Spongebob
2:04 Interesting PS Generative Fill
2:09 QR Code ControlNET
2:19 White House vs Safe.ai
2:39 Safe.ai vs Andrew Ng
3:09 SAM-HQ
3:45 T2V Potat1
3:57 smart?
4:05 Air Force Fake AI News
4:14 StyleDrop
4:32 Chatbots = Humans Downfall?
5:18 Japan AI Copyright Law (& Israel)
5:42 RIP OpenAI CTO Twitter
5:51 Hey ChatGPT, finish this building...
5:57 Time Magazine New AI Cover
6:02 OpenAI Cybersecurity Grant
6:08 Snap Fusion
6:15 Bytes Are All You Need
6:57 OpenAI Got Better At Maths
7:05 Redream
7:18 GPTFile
7:22 Runway Gen2
7:28 Outro")
![Researchers Are Getting Really Creative Training LLMs [Token Order Prediction]](https://i.ytimg.com/vi/sgIB7l6hW3Q/mqdefault.jpg "Researchers Are Getting Really Creative Training LLMs [Token Order Prediction]
Deploy on Sevalla now and get a free $50 credit!
https://sevalla.com/?utm_source=ByCloud%27s&utm_medium=Referral&utm_campaign=youtube
Meta’s 2024 paper explores Multi-Token Prediction (MTP), where LLMs predict several future tokens at once to improve efficiency and its foresight. With DeepSeek 2024 then proposing it as an auxiliary objective, a cool concept to improve model performance during training. But MTP may not be the ideal AO yet. In this video, Ill be introducing Token Order Prediction (TOP), a softer, auxiliary objective that enhances model performance by learning the relative order of upcoming tokens.
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
Multi-token Prediction
[Paper] https://arxiv.org/abs/2404.19737
Token Order Prediction
[Paper] https://arxiv.org/abs/2508.19228
DeepSeek V3
[Paper] https://arxiv.org/abs/2412.19437v1
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")
