Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

Axiom Math founder Carina Hong on why math is the path to general superintelligence: “Math is the sandbox for reality. You very quickly see verifiable rewards because in math, there’s an absolute right or wrong.” “And especially when you have Lean, you can check the proof or solution step... show more

TBPN

586,135 subscribers

39,589 görüntüleme • 4 ay önce •via X (Twitter)

Bilim & Teknoloji

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

“Math is code. And code is math.” Carina Hong CEO Axiom on why that idea matters for AI: “For centuries, mathematicians reasoned in natural language. But because of the Curry–Howard correspondence, a proof can also be a program.” “You can translate natural-language math into formal code using systems like Lean. Lean is both a theorem-proving language and a programming language.” “That means you can generate solutions as code and then formally verify them. You can even use the same system to verify programs themselves.” “This creates a flywheel of generation and verification, where AI can both write solutions and prove that they’re correct.”

“Math is code. And code is math.” Carina Hong CEO Axiom on why that idea matters for AI: “For centuries, mathematicians reasoned in natural language. But because of the Curry–Howard correspondence, a proof can also be a program.” “You can translate natural-language math into formal code using systems like Lean. Lean is both a theorem-proving language and a programming language.” “That means you can generate solutions as code and then formally verify them. You can even use the same system to verify programs themselves.” “This creates a flywheel of generation and verification, where AI can both write solutions and prove that they’re correct.”

Forward Future

24,729 görüntüleme • 5 ay önce

Terence Tao summarized how AI is massively accelerating math career and math research. "In math, you previously had to basically go through years and years of education to be a math PhD before you could contribute to the frontier of math research. But now it's quite possible at the high school level or whatever, that you could get involved in a math project and actually make a real contribution because of all these AI tools and lean and everything else." From Dwarkesh Patel podcast (link to full video in comment)

Terence Tao summarized how AI is massively accelerating math career and math research. "In math, you previously had to basically go through years and years of education to be a math PhD before you could contribute to the frontier of math research. But now it's quite possible at the high school level or whatever, that you could get involved in a math project and actually make a real contribution because of all these AI tools and lean and everything else." From Dwarkesh Patel podcast (link to full video in comment)

Rohan Paul

61,705 görüntüleme • 2 ay önce

Why is AI suddenly cracking all kinds of previously unsolved problems in math? My guess is that it’s because: A) Math is verifiable so it's easier for models to improve B) Theoretical math highly benefits from chain of thought reasoning C) The answers and breakthroughs can be tested and verified by expert humans. But this leads to another question: what if progress in other fields is moving just as fast and we can't see it?

Why is AI suddenly cracking all kinds of previously unsolved problems in math? My guess is that it’s because: A) Math is verifiable so it's easier for models to improve B) Theoretical math highly benefits from chain of thought reasoning C) The answers and breakthroughs can be tested and verified by expert humans. But this leads to another question: what if progress in other fields is moving just as fast and we can't see it?

nxthompson

12,780 görüntüleme • 4 gün önce

Why do we do math in school? When I asked this question as a kid, my teachers always told me, "You'll use it one day." That's almost always false. Most adults are never going to be in a sticky situation with two binomials, thinking, thank goodness I can FOIL my way out of this mess. Most of us get by just fine without ever solving for x again. And especially in an Al age, fewer and fewer people will "use math" in the narrow, practical sense. But the origins of math have never been primarily about utility. Math is formative. It trains the mind to love what is true, to recognize what is orderly, and to be drawn toward what is beautiful. It teaches us that the universe is not chaos, but something intelligible, something structured, something that can be known. And you can see that truth made visible in the world's most breathtaking churches and cathedrals: arches, vaults, domes, proportions, symmetry, harmony, light-geometry turned into glory. The mathematics that shapes a cathedral is not cold or sterile. It's the language of wonder, carved into stone. We don't teach math because everyone will use it. We teach math because it forms the kind of person who can see that reality has meaning.

Why do we do math in school? When I asked this question as a kid, my teachers always told me, "You'll use it one day." That's almost always false. Most adults are never going to be in a sticky situation with two binomials, thinking, thank goodness I can FOIL my way out of this mess. Most of us get by just fine without ever solving for x again. And especially in an Al age, fewer and fewer people will "use math" in the narrow, practical sense. But the origins of math have never been primarily about utility. Math is formative. It trains the mind to love what is true, to recognize what is orderly, and to be drawn toward what is beautiful. It teaches us that the universe is not chaos, but something intelligible, something structured, something that can be known. And you can see that truth made visible in the world's most breathtaking churches and cathedrals: arches, vaults, domes, proportions, symmetry, harmony, light-geometry turned into glory. The mathematics that shapes a cathedral is not cold or sterile. It's the language of wonder, carved into stone. We don't teach math because everyone will use it. We teach math because it forms the kind of person who can see that reality has meaning.

Jeremy Wayne Tate

95,173 görüntüleme • 6 ay önce

Common Core math was introduced in 2010 to address standards across the U.S. and shift the focus from rote memorization to a deeper, conceptual understanding of mathematics. By 2026, it remains a common standard in many states, aiming to prepare students for college and the modern workforce by fostering critical thinking and number sense rather than just memorizing formulas. There has been much debate on using common core math in schools curriculum. Do you agree with using common core math in schools? Or do you think the old way of doing math was the better way to teach math to children?

Common Core math was introduced in 2010 to address standards across the U.S. and shift the focus from rote memorization to a deeper, conceptual understanding of mathematics. By 2026, it remains a common standard in many states, aiming to prepare students for college and the modern workforce by fostering critical thinking and number sense rather than just memorizing formulas. There has been much debate on using common core math in schools curriculum. Do you agree with using common core math in schools? Or do you think the old way of doing math was the better way to teach math to children?

🌸𝓐𝓾𝓭𝓻𝓮𝔂🌸

60,861 görüntüleme • 2 ay önce

American mother says her daughter is 1 of only 4 children in her entire math class that speaks English “She doesn't get to learn math because there's 20 plus other kids who don't know English and the teacher has to accommodate them” “Why am I paying property taxes? Why am I paying income tax? Why am I paying taxes for anything for my children to have a the right to a free and public education? — This this is what they get. You don't get to learn math because you know English, so you just don't get to learn math.”

American mother says her daughter is 1 of only 4 children in her entire math class that speaks English “She doesn't get to learn math because there's 20 plus other kids who don't know English and the teacher has to accommodate them” “Why am I paying property taxes? Why am I paying income tax? Why am I paying taxes for anything for my children to have a the right to a free and public education? — This this is what they get. You don't get to learn math because you know English, so you just don't get to learn math.”

Wall Street Apes

2,293,351 görüntüleme • 6 ay önce

Terence Tao: "Previously, you needed a PhD to contribute to math research. Now a high school student can." Dwarkesh asks the world's most famous mathematician: what's your advice for someone considering a career in math, especially in light of AI progress? Tao is honest about uncertainty: "We live in a time of change. A particularly unpredictable era. Things that we've taken for granted for centuries may not hold anymore. The way we do everything... not just mathematics... will change." He admits his preference: "In many ways, I would prefer a much more boring, quiet era where things are much the same as they were 10 or 20 years ago. But one just has to embrace this. There's going to be a lot of change. The things you study... some of them may become obsolete or revolutionized. But some things will be retained." On new opportunities: "Previously, you had to go through years and years of education and get a math PhD before you could contribute to the frontier of math research. But now it's quite possible at the high school level that you could get involved in a math project and actually make a real contribution... because of all these AI tools and Lean and everything else." His advice: "There will be a lot of non-traditional opportunities to learn. You need a very adaptable mindset. There'll be worth pursuing things just for curiosity and for playing around. Still go through traditional education and learn math and science the old-fashioned way for a while... credentials will still be important. But you should also be open to very, very different ways of doing science. Some of which don't exist yet." He concludes: "It's a scary time. But also very exciting."

Terence Tao: "Previously, you needed a PhD to contribute to math research. Now a high school student can." Dwarkesh asks the world's most famous mathematician: what's your advice for someone considering a career in math, especially in light of AI progress? Tao is honest about uncertainty: "We live in a time of change. A particularly unpredictable era. Things that we've taken for granted for centuries may not hold anymore. The way we do everything... not just mathematics... will change." He admits his preference: "In many ways, I would prefer a much more boring, quiet era where things are much the same as they were 10 or 20 years ago. But one just has to embrace this. There's going to be a lot of change. The things you study... some of them may become obsolete or revolutionized. But some things will be retained." On new opportunities: "Previously, you had to go through years and years of education and get a math PhD before you could contribute to the frontier of math research. But now it's quite possible at the high school level that you could get involved in a math project and actually make a real contribution... because of all these AI tools and Lean and everything else." His advice: "There will be a lot of non-traditional opportunities to learn. You need a very adaptable mindset. There'll be worth pursuing things just for curiosity and for playing around. Still go through traditional education and learn math and science the old-fashioned way for a while... credentials will still be important. But you should also be open to very, very different ways of doing science. Some of which don't exist yet." He concludes: "It's a scary time. But also very exciting."

Jaynit

77,410 görüntüleme • 3 ay önce

A Canadian math education association put “2 + 2 = 4” on a white supremacy slide. In 2022, the Ontario Mathematics Coordinators Association ran a professional development webinar for math teachers across Ontario school boards. The slide was titled “White Supremacy in Math Education.” Under “covert white supremacy” it listed: standardized testing, “I don’t see colour in my math class,” and “Of course math is neutral because 2 + 2 = 4.” Not a fringe blog. Not a parody account. A professional association shaping how math teachers are trained. When the slide went viral in 2023, the OMCA released an official statement confirming that 2 + 2 = 4 is mathematically true — but argued the equation can be “weaponized to uphold bigotry and white supremacy ideology” when used to claim math is politically neutral. Their defense of the slide was the slide. Someone once wrote about a party that demanded citizens believe two and two made five. In 2022, a professional math association wrote a slide that said believing math is neutral is white supremacy. One was fiction. The other was professional development.

A Canadian math education association put “2 + 2 = 4” on a white supremacy slide. In 2022, the Ontario Mathematics Coordinators Association ran a professional development webinar for math teachers across Ontario school boards. The slide was titled “White Supremacy in Math Education.” Under “covert white supremacy” it listed: standardized testing, “I don’t see colour in my math class,” and “Of course math is neutral because 2 + 2 = 4.” Not a fringe blog. Not a parody account. A professional association shaping how math teachers are trained. When the slide went viral in 2023, the OMCA released an official statement confirming that 2 + 2 = 4 is mathematically true — but argued the equation can be “weaponized to uphold bigotry and white supremacy ideology” when used to claim math is politically neutral. Their defense of the slide was the slide. Someone once wrote about a party that demanded citizens believe two and two made five. In 2022, a professional math association wrote a slide that said believing math is neutral is white supremacy. One was fiction. The other was professional development.

Jake

112,641 görüntüleme • 4 ay önce

I had a fantastic time discussing with the learning legend Justin Skycak from Math Academy about learning math in the modern age. we've talked about his quite impressive self-learning journey (3000h of math in high school) all the way to how he hand curated the initial knowledge graph for math academy to make that process more efficient. great lively 3h discussion here are the chapters: 0:00:00 - intro: 0:02:10 - justin background 0:05:45 - 3000h math self study in high school 0:11:45 - what a day looked like for that 3000h stretch 0:16:10 - meta-learning vs pure math learning 0:21:50 - when did you get into cognitive neuro? 0:29:55 - how did the fundamental math helped in your research projects 0:43:10 - what does the math academy learning system looks like 0:47:34 - how did you guys build the 2000 topic knowledge graph 1:01:15 - would LLM be useful as an interface to that knowledge graph for the students? 1:10:46 - how does the FIRe spaced repetition algorithm works? 1:17:34 - does the same knowledge graph structure would work for physics? or other topic?: 1:34:05 - how do you understand the subject vs the curiculum 1:35:50 - is there a connection between studying math and learning a sport? 1:42:00 - do you think in math doing and teaching requires different skills? 1:56:25 - could you get understanding without automaticy? 2:05:35 - do you see any upside of confusion in learning? 2:14:11 - learning math as an adult? 2:19:20 - how to fill the motivation gap after learning the fundamental? 2:24:10 - how should teaching math for kids and adults balance fundamentals and creativity? 2:33:55 - is it ever too late to learn math seriously? 2:46:00 - mastery learning vs ultra learning 2:51:30 - top-down vs bottom-up 2:53:40 - mastery learning for domain without a structured hierarchical structure? 2:56:30 - neurodivergence / adhd for structured math learning? 3:06:20 - amateur mathematician augmented with technology will be able to contribute to research? 3:14:37 - what are you most excited about right now in term of learning enjoy!

I had a fantastic time discussing with the learning legend Justin Skycak from Math Academy about learning math in the modern age. we've talked about his quite impressive self-learning journey (3000h of math in high school) all the way to how he hand curated the initial knowledge graph for math academy to make that process more efficient. great lively 3h discussion here are the chapters: 0:00:00 - intro: 0:02:10 - justin background 0:05:45 - 3000h math self study in high school 0:11:45 - what a day looked like for that 3000h stretch 0:16:10 - meta-learning vs pure math learning 0:21:50 - when did you get into cognitive neuro? 0:29:55 - how did the fundamental math helped in your research projects 0:43:10 - what does the math academy learning system looks like 0:47:34 - how did you guys build the 2000 topic knowledge graph 1:01:15 - would LLM be useful as an interface to that knowledge graph for the students? 1:10:46 - how does the FIRe spaced repetition algorithm works? 1:17:34 - does the same knowledge graph structure would work for physics? or other topic?: 1:34:05 - how do you understand the subject vs the curiculum 1:35:50 - is there a connection between studying math and learning a sport? 1:42:00 - do you think in math doing and teaching requires different skills? 1:56:25 - could you get understanding without automaticy? 2:05:35 - do you see any upside of confusion in learning? 2:14:11 - learning math as an adult? 2:19:20 - how to fill the motivation gap after learning the fundamental? 2:24:10 - how should teaching math for kids and adults balance fundamentals and creativity? 2:33:55 - is it ever too late to learn math seriously? 2:46:00 - mastery learning vs ultra learning 2:51:30 - top-down vs bottom-up 2:53:40 - mastery learning for domain without a structured hierarchical structure? 2:56:30 - neurodivergence / adhd for structured math learning? 3:06:20 - amateur mathematician augmented with technology will be able to contribute to research? 3:14:37 - what are you most excited about right now in term of learning enjoy!

Yacine Mahdid

57,320 görüntüleme • 3 ay önce

This is a day in the life of a Flat Earther You tell us your math We show you real life examples of how your math doesn’t fit reality You throw in some excuses that don’t explain anything while simultaneously refusing to actually look at the evidence Because science dudes! 😎

This is a day in the life of a Flat Earther You tell us your math We show you real life examples of how your math doesn’t fit reality You throw in some excuses that don’t explain anything while simultaneously refusing to actually look at the evidence Because science dudes! 😎

Ted Logan

89,127 görüntüleme • 3 yıl önce

Terence Tao thinks AI is already very good at using existing, well-understood math techniques to solve problems. An important question is how many open problems in math could be solved this way, without developing any new ideas. An extreme case of a proof like this is the four-color theorem, which was proved by checking a thousand examples one-by-one. If there are lots of problems that are vulnerable to just mechanically applying existing techniques, AI could start to produce new math results incredibly rapidly. But these proofs might not help advance our civilization’s knowledge the way math has in the past.

Terence Tao thinks AI is already very good at using existing, well-understood math techniques to solve problems. An important question is how many open problems in math could be solved this way, without developing any new ideas. An extreme case of a proof like this is the four-color theorem, which was proved by checking a thousand examples one-by-one. If there are lots of problems that are vulnerable to just mechanically applying existing techniques, AI could start to produce new math results incredibly rapidly. But these proofs might not help advance our civilization’s knowledge the way math has in the past.

Dwarkesh Patel

64,547 görüntüleme • 4 ay önce

Terence Tao: AI isn’t hype anymore in Math discovery. Terence Tao is one of the greatest living mathematicians, in his new lecture explains how AI and human professional mathematicians are now complementary. "There has been a really visible increase in capability. It is not pure hype by any means. To me, these advances show there is a complementary way to do mathematics. Humans traditionally work in small groups on hard problems for months, and we will keep doing that. But we can also now set AI to scale: sweep a thousand problems and pick up all the low-hanging fruit. Figure out all the ways to match problems to methods. If there are 20 different techniques, apply them all to 1,000 problems and see which ones can be solved by these methods. This is the capability that is present today." From 'Institute for Pure & Applied Mathematics (IPAM)' YT channel.

Terence Tao: AI isn’t hype anymore in Math discovery. Terence Tao is one of the greatest living mathematicians, in his new lecture explains how AI and human professional mathematicians are now complementary. "There has been a really visible increase in capability. It is not pure hype by any means. To me, these advances show there is a complementary way to do mathematics. Humans traditionally work in small groups on hard problems for months, and we will keep doing that. But we can also now set AI to scale: sweep a thousand problems and pick up all the low-hanging fruit. Figure out all the ways to match problems to methods. If there are 20 different techniques, apply them all to 1,000 problems and see which ones can be solved by these methods. This is the capability that is present today." From 'Institute for Pure & Applied Mathematics (IPAM)' YT channel.

Rohan Paul

174,087 görüntüleme • 5 ay önce

The way math is presented can be really misleading about the actual nature of math. In theory, someone could have discovered quantum information theory decades before quantum mechanics even existed. It's just about combining linear algebra with probability. But nobody would have known it was worth combining these fields - or how, exactly, to combine them - before QM. Math papers and textbooks go rigourously step-by-step, starting with axioms and definitions, then deducing theorems. It's great for making sure proofs can be verified and ensuring the subject is based on sound logic. But for learners, it hides the really important thing: where did the axioms come from? Why these definitions and not others? The thing that's missing is the sense that there were concrete problems that these ideas helped solve.

The way math is presented can be really misleading about the actual nature of math. In theory, someone could have discovered quantum information theory decades before quantum mechanics even existed. It's just about combining linear algebra with probability. But nobody would have known it was worth combining these fields - or how, exactly, to combine them - before QM. Math papers and textbooks go rigourously step-by-step, starting with axioms and definitions, then deducing theorems. It's great for making sure proofs can be verified and ensuring the subject is based on sound logic. But for learners, it hides the really important thing: where did the axioms come from? Why these definitions and not others? The thing that's missing is the sense that there were concrete problems that these ideas helped solve.

Dwarkesh Patel

58,236 görüntüleme • 1 ay önce

When you apply the math and their claimed dimensions for the globe and compare it to reality…. You’re left with two choices! You can ignore it and continue believing in the fairytale they have provided for you. Or you can accept the truth that our world is NOT what they have told us. After all, who are you gonna believe, your loving government or your lying eyes?!?!

When you apply the math and their claimed dimensions for the globe and compare it to reality…. You’re left with two choices! You can ignore it and continue believing in the fairytale they have provided for you. Or you can accept the truth that our world is NOT what they have told us. After all, who are you gonna believe, your loving government or your lying eyes?!?!

Ted Theodore Logan

18,984 görüntüleme • 1 yıl önce

Thiel Fellow and cofounder of The Antifraud Company Alex Shieh says he's building the Palantir of fraud detection: "People have always been able to find fraud manually. But usually they either luck into it or they have a tip from an insider. "We're trying to eliminate that bottleneck. We want to find the fraud using AI." "We're tapping into all these sources of data. And once we have a good idea of what's going on, we send out human investigators to go talk to sources." "But we want to be going outbound. We don't want to have to wait for people to come to us. Because that's just less efficient. When there's $600 billion in fraud, you're not going to get tips for most of it. And if you want to solve it, this is the way you have to go."

Thiel Fellow and cofounder of The Antifraud Company Alex Shieh says he's building the Palantir of fraud detection: "People have always been able to find fraud manually. But usually they either luck into it or they have a tip from an insider. "We're trying to eliminate that bottleneck. We want to find the fraud using AI." "We're tapping into all these sources of data. And once we have a good idea of what's going on, we send out human investigators to go talk to sources." "But we want to be going outbound. We don't want to have to wait for people to come to us. Because that's just less efficient. When there's $600 billion in fraud, you're not going to get tips for most of it. And if you want to solve it, this is the way you have to go."

TBPN

21,560 görüntüleme • 3 ay önce

How does math research change when the cost of trying your first dumb idea goes to zero? Daniel Litt joins Greg Burnham and Anson Ho to discuss what today’s models can and can’t do in math, and how far they are from doing high-quality research. 0:00:00 What's the hardest math problem AI can solve today? 00:16:08 How helpful are today’s AI models for math research? 00:23:36 Junk papers, LLM-generated proofs, and the refereeing crisis 00:27:21 AI enables searching through problems at scale 00:33:49 When will AI be good enough to publish in top math journals? 00:42:15 What are the returns to intelligence? 00:59:50 Will AI solve Millennium problems? 01:11:54 Is math full of low-hanging fruit? 01:18:47 How Daniel has adapted his professional life to AI progress 01:25:28 What do AI math benchmarks actually measure? 01:33:05 Designing the Open Problems benchmark 01:56:35 Do mathematicians believe heuristic arguments about conjectures? 02:01:24 What if FrontierMath: Open Problems gets solved? 02:06:53 Is AI on the cusp of accelerating math progress?

How does math research change when the cost of trying your first dumb idea goes to zero? Daniel Litt joins Greg Burnham and Anson Ho to discuss what today’s models can and can’t do in math, and how far they are from doing high-quality research. 0:00:00 What's the hardest math problem AI can solve today? 00:16:08 How helpful are today’s AI models for math research? 00:23:36 Junk papers, LLM-generated proofs, and the refereeing crisis 00:27:21 AI enables searching through problems at scale 00:33:49 When will AI be good enough to publish in top math journals? 00:42:15 What are the returns to intelligence? 00:59:50 Will AI solve Millennium problems? 01:11:54 Is math full of low-hanging fruit? 01:18:47 How Daniel has adapted his professional life to AI progress 01:25:28 What do AI math benchmarks actually measure? 01:33:05 Designing the Open Problems benchmark 01:56:35 Do mathematicians believe heuristic arguments about conjectures? 02:01:24 What if FrontierMath: Open Problems gets solved? 02:06:53 Is AI on the cusp of accelerating math progress?

Epoch AI

178,703 görüntüleme • 6 ay önce

New Course: Reinforcement Fine-Tuning LLMs with GRPO! Learn to use reinforcement learning to improve your LLM performance in this short course, built in collaboration with Predibase by Rubrik, and taught by Travis Addair, its Co-Founder and CTO, and Arnav Garg, its Senior Engineer and Machine Learning Lead. Reasoning models have been one of the most important developments in LLMs. Reinforcement Fine-Tuning (RFT) uses rewards to encourage LLMs to find solutions to multi-step reasoning tasks such as solving math problems and debugging code - without needing pre-existing training examples like in traditional supervised fine-tuning. Group Relative Policy Optimization (GRPO) is a reinforcement fine-tuning algorithm gaining rapid adoption. Developed by the DeepSeek team and used to train the R1 reasoning model, GRPO uses reward functions that you can write in Python to assign rewards to model responses. It’s beneficial for tasks with verifiable outcomes and can work well even with fewer than 100 training examples. It can also significantly improve the reasoning ability of smaller LLMs, making applications faster and more cost effective. In this course, you’ll take a technical deep dive into RFT with GRPO. You’ll learn to build reward functions that you can use in the GRPO training process to guide an LLM toward better performance on multi-step reasoning tasks. In detail, you’ll: - Learn when reinforcement fine-tuning is a better fit than supervised fine-tuning, especially for tasks involving multi-step reasoning or limited labeled data. - Understand how GRPO uses programmable reward functions as a more scalable alternative to the human feedback required for other reinforcement learning algorithms, such as RLHF and DPO. - Frame the Wordle game as a reinforcement fine-tuning problem and see how an LLM can learn to plan, analyze feedback, and improve its strategy over time. - Design reward functions that power the reinforcement fine-tuning process. - Learn techniques for evaluating more subjective tasks, such as rating the quality of a text summary, using an LLM as a judge. - Understand why reward hacking happens and how to avoid it by adding penalty functions to discourage undesirable behaviors. - Learn the four key components of the loss calculation in the GRPO algorithm: token probability distribution ratios, advantages, clipping, and KL-divergence. - Launch reinforcement fine-tuning jobs using Predibase’s hosted training services. By the end of this course, you’ll be able to build and fine-tune LLMs using reinforcement learning to improve reasoning without relying on large labeled datasets or subjective human feedback. Please sign up here:

New Course: Reinforcement Fine-Tuning LLMs with GRPO! Learn to use reinforcement learning to improve your LLM performance in this short course, built in collaboration with Predibase by Rubrik, and taught by Travis Addair, its Co-Founder and CTO, and Arnav Garg, its Senior Engineer and Machine Learning Lead. Reasoning models have been one of the most important developments in LLMs. Reinforcement Fine-Tuning (RFT) uses rewards to encourage LLMs to find solutions to multi-step reasoning tasks such as solving math problems and debugging code - without needing pre-existing training examples like in traditional supervised fine-tuning. Group Relative Policy Optimization (GRPO) is a reinforcement fine-tuning algorithm gaining rapid adoption. Developed by the DeepSeek team and used to train the R1 reasoning model, GRPO uses reward functions that you can write in Python to assign rewards to model responses. It’s beneficial for tasks with verifiable outcomes and can work well even with fewer than 100 training examples. It can also significantly improve the reasoning ability of smaller LLMs, making applications faster and more cost effective. In this course, you’ll take a technical deep dive into RFT with GRPO. You’ll learn to build reward functions that you can use in the GRPO training process to guide an LLM toward better performance on multi-step reasoning tasks. In detail, you’ll: - Learn when reinforcement fine-tuning is a better fit than supervised fine-tuning, especially for tasks involving multi-step reasoning or limited labeled data. - Understand how GRPO uses programmable reward functions as a more scalable alternative to the human feedback required for other reinforcement learning algorithms, such as RLHF and DPO. - Frame the Wordle game as a reinforcement fine-tuning problem and see how an LLM can learn to plan, analyze feedback, and improve its strategy over time. - Design reward functions that power the reinforcement fine-tuning process. - Learn techniques for evaluating more subjective tasks, such as rating the quality of a text summary, using an LLM as a judge. - Understand why reward hacking happens and how to avoid it by adding penalty functions to discourage undesirable behaviors. - Learn the four key components of the loss calculation in the GRPO algorithm: token probability distribution ratios, advantages, clipping, and KL-divergence. - Launch reinforcement fine-tuning jobs using Predibase’s hosted training services. By the end of this course, you’ll be able to build and fine-tune LLMs using reinforcement learning to improve reasoning without relying on large labeled datasets or subjective human feedback. Please sign up here:

Andrew Ng

86,457 görüntüleme • 1 yıl önce

This becomes a math problem really quickly. If Iran is able to keep firing missiles and drones, we could run low on interceptors. The administration insists they have enough for the mission, but they also can't give a straight answer about how long this is going to last. That could put more U.S. servicemembers at risk.

This becomes a math problem really quickly. If Iran is able to keep firing missiles and drones, we could run low on interceptors. The administration insists they have enough for the mission, but they also can't give a straight answer about how long this is going to last. That could put more U.S. servicemembers at risk.

Senator Mark Kelly

53,312 görüntüleme • 5 ay önce

IN NEWS: Attio raises $52M series B led by GV. Nicolas Sharp on AI in CRM: "LLMs work great when they have lots of context and you need to feed them that. One of the first things we worked on was this data model and being able to instantly ingest emails, calendars, product data, etc." "We have a thesis that code is becoming the new no-code. Software becomes so much more malleable because code is so much easier to generate."

IN NEWS: Attio raises $52M series B led by GV. Nicolas Sharp on AI in CRM: "LLMs work great when they have lots of context and you need to feed them that. One of the first things we worked on was this data model and being able to instantly ingest emails, calendars, product data, etc." "We have a thesis that code is becoming the new no-code. Software becomes so much more malleable because code is so much easier to generate."

TBPN

16,200 görüntüleme • 11 ay önce

Atreides' Gavin Baker says that once Starship is reusable, the economics of orbital compute crush the economics of terrestrial compute. He lays out the math: "Once $SPCX can reuse Starship, the math for orbital compute becomes pretty compelling." "It's $60B to bring on a GW terrestrially. $25B is power and cooling — you don't need that in space. And so the right comp for that $35B in IT equipment — GPU, CPU, switches, memory, storage — is that $25B [in power and cooling] vs. the cost of launch." "And once Starship is reusable, I think the cost to launch is $5B. So that means you can put a GW into space for [$40B]. The GW on earth is $60B."

Atreides' Gavin Baker says that once Starship is reusable, the economics of orbital compute crush the economics of terrestrial compute. He lays out the math: "Once $SPCX can reuse Starship, the math for orbital compute becomes pretty compelling." "It's $60B to bring on a GW terrestrially. $25B is power and cooling — you don't need that in space. And so the right comp for that $35B in IT equipment — GPU, CPU, switches, memory, storage — is that $25B [in power and cooling] vs. the cost of launch." "And once Starship is reusable, I think the cost to launch is $5B. So that means you can put a GW into space for [$40B]. The GW on earth is $60B."

TBPN

102,778 görüntüleme • 1 ay önce