![Researchers Are Getting Really Creative Training LLMs [Token Order Prediction] @bycloudAI](https://i.ytimg.com/vi/sgIB7l6hW3Q/maxresdefault.jpg)
Uploaded October 2025 | Updated September 2026, 2 weeks ago
Deploy on Sevalla now and get a free $50 credit!
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, I'll be introducing Token Order Prediction (TOP), a softer, auxiliary objective that enhances model performance by learning the relative order of upcoming tokens.
My Newsletter
mail.bycloud.ai
my project: find, discover & explain AI research semantically
findmypapers.ai
My Patreon
patreon.com/c/bycloud
Multi-token Prediction
[Paper] arxiv.org/abs/2404.19737
Token Order Prediction
[Paper] arxiv.org/abs/2508.19228
DeepSeek V3
[Paper] arxiv.org/abs/2412.19437v1
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
[Discord] discord.gg/NhJZGtH
[Twitter] twitter.com/bycloudai
[Patreon] patreon.com/bycloud
[Business Inquiries] bycloud@smoothmedia.co
[Profile & Banner Art] twitter.com/pygm7
[Video Editor] @Booga04
[Ko-fi] ko-fi.com/bycloudai
Deploy on Sevalla now and get a free $50 credit!
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, I'll be introducing Token Order Prediction (TOP), a softer, auxiliary objective that enhances model performance by learning the relative order of upcoming tokens.
My Newsletter
mail.bycloud.ai
my project: find, discover & explain AI research semantically
findmypapers.ai
My Patreon
patreon.com/c/bycloud
Multi-token Prediction
[Paper] arxiv.org/abs/2404.19737
Token Order Prediction
[Paper] arxiv.org/abs/2508.19228
DeepSeek V3
[Paper] arxiv.org/abs/2412.19437v1
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
[Discord] discord.gg/NhJZGtH
[Twitter] twitter.com/bycloudai
[Patreon] patreon.com/bycloud
[Business Inquiries] bycloud@smoothmedia.co
[Profile & Banner Art] twitter.com/pygm7
[Video Editor] @Booga04
[Ko-fi] ko-fi.com/bycloudai
![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")
![If AIs So Good, Why Cant It Make Its Own Data? [SEAL]](https://i.ytimg.com/vi/uOO02o4J5KU/mqdefault.jpg "If AIs So Good, Why Cant It Make Its Own Data? [SEAL]
Get started now with privacy focused VPN by Proton! https://proton.me/pass/bycloudai
Can a neural network write its own data and skyrocket past GPT-4? In todays video, we dissect the brand-new “Self-Adapting Language Models” paper (SEAL), where an LLM fabricates synthetic data, tunes LoRA adapters, and after just two rounds, outperforms much larger models on SQuAD and ARC.
Self-Adapting Language Models
[Paper] https://arxiv.org/abs/2506.10943
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
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:
🙏Nous Research, Chris LeDoux, Ben Shaener, DX Research Group, Poof N Inu, Andrew Lescelius, 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] Abhay
[Ko-fi] https://ko-fi.com/bycloudai")
![New AI Meta: Train LLMs To Explore On Hard Tokens [RLVR + Entropy]](https://i.ytimg.com/vi/uOrJUksvIhs/mqdefault.jpg "New AI Meta: Train LLMs To Explore On Hard Tokens [RLVR + Entropy]
Get started with Strands Agents today:
https://aws.amazon.com/blogs/opensource/introducing-strands-agents-1-0-production-ready-multi-agent-orchestration-made-simple/?trk=3fa00bba-d2bc-45e2-a2b0-f96bc12fd521&sc_channel=psm
In this video, I will be sharing how researchers train LLMs to explore during RL to improve performance via entropy.
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
Beyond the 80/20 Rule
[Paper] https://arxiv.org/abs/2506.01939
Reasoning with Exploration
[Paper] https://arxiv.org/abs/2506.14758
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:
🙏Nous Research, Chris LeDoux, Ben Shaener, DX Research Group, Poof N Inu, Andrew Lescelius, 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")
