Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

Amazing accuracy.

The Figen

2,773,803 subscribers

11,052,783 просмотров • 2 лет назад •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

Комментарии: 10

Фото профиля 💪🎭..Rai ji..💪🎭

💪🎭..Rai ji..💪🎭2 лет назад

Huge respect to her. An inspiration ❤️

Фото профиля Amit Shah (Parody)

Amit Shah (Parody)2 лет назад

Goals like poetry.

Фото профиля 💪🎭..Rai ji..💪🎭

💪🎭..Rai ji..💪🎭2 лет назад

In 1995, an 8-year-old boy appeared playing in a tournament in Rosario, Argentina. His name was Lionel Messi.

Фото профиля CHIKU JI❤️💫

CHIKU JI❤️💫2 лет назад

Humanity ❤️

Фото профиля Amit Shah (Parody)

Amit Shah (Parody)2 лет назад

One of the best goals in football history ...

Фото профиля Wholesome Side of 𝕏

Wholesome Side of 𝕏2 лет назад

He is so happy after scoring that goal 🤗

Фото профиля 💪🎭..Rai ji..💪🎭

💪🎭..Rai ji..💪🎭2 лет назад

Respect 👏

Фото профиля G.r.a.c.h.i. M.e.m.e

G.r.a.c.h.i. M.e.m.e2 лет назад

Accurate too🌚

Фото профиля Precious ❤️ 🇳🇬

Precious ❤️ 🇳🇬2 лет назад

Not possible

Фото профиля Christian C.🌷

Christian C.🌷2 лет назад

This where I lost my breath🫡

Похожие видео

Accuracy in different fields is amazing of talented ones.

Accuracy in different fields is amazing of talented ones.

BuaaGlowVibes

3,472,826 просмотров • 1 месяц назад

I have missed my favorite Kung Fu couple. The accuracy in their moves with the sound is absolutely amazing

I have missed my favorite Kung Fu couple. The accuracy in their moves with the sound is absolutely amazing

Dr. God Abeg ooo

66,689 просмотров • 7 месяцев назад

AMAZING VIDEO: Richard Nixon predicted Putinism and the revival of Russian Imperialism with stunning accuracy in this 1994 interview

AMAZING VIDEO: Richard Nixon predicted Putinism and the revival of Russian Imperialism with stunning accuracy in this 1994 interview

Business Ukraine mag

656,017 просмотров • 2 лет назад

Amazing how clean and precise this snake's movement is. She follows a pattern, skipping one hole and slipping through the next with remarkable accuracy. 📹 : Liam Zdarec

Amazing how clean and precise this snake's movement is. She follows a pattern, skipping one hole and slipping through the next with remarkable accuracy. 📹 : Liam Zdarec

Interesting Universe

194,523 просмотров • 1 месяц назад

CENTER MAGIC JOHNSON BUILD is AMAZING on NBA 2K26! 92 Close Shot - 98 Layup 87 Mid - 75 3pt 99 Pass Accuracy - 80 Ball Handle 92 Rebound and more... Trying to get this video out tomorrow!

CENTER MAGIC JOHNSON BUILD is AMAZING on NBA 2K26! 92 Close Shot - 98 Layup 87 Mid - 75 3pt 99 Pass Accuracy - 80 Ball Handle 92 Rebound and more... Trying to get this video out tomorrow!

✞ S̴̴p̴̴r̴̴a̴̴g̴̴u̴̴e̴̴2̴̴k̴ ✞

42,207 просмотров • 5 месяцев назад

🔥 Shedeur & Pro-Bowl. Denzel Ward High Praise for #12 💛 "I think he has all the confidence in the world. He was making amazing throws, has great accuracy, and poised in there. Time under task"

🔥 Shedeur & Pro-Bowl. Denzel Ward High Praise for #12 💛 "I think he has all the confidence in the world. He was making amazing throws, has great accuracy, and poised in there. Time under task"

JaKi 🇺🇸

56,964 просмотров • 7 месяцев назад

LeTHANOS JAMES BUILD is AMAZING on NBA 2K26! W/ CAPS & MORE: 94 DRIVING DUNK 85 LAYUP 85 STANDING DUNK 90 3PT= LeFuckYou TAKEOVER 92 PASS ACCURACY 92 PERIMETER 80 REBOUND 85 SPEED/83 AGILITY=SUGGS MOTION STYLE 86 BALL HANDLE 79 POST CONTROL

LeTHANOS JAMES BUILD is AMAZING on NBA 2K26! W/ CAPS & MORE: 94 DRIVING DUNK 85 LAYUP 85 STANDING DUNK 90 3PT= LeFuckYou TAKEOVER 92 PASS ACCURACY 92 PERIMETER 80 REBOUND 85 SPEED/83 AGILITY=SUGGS MOTION STYLE 86 BALL HANDLE 79 POST CONTROL

✞ S̴̴p̴̴r̴̴a̴̴g̴̴u̴̴e̴̴2̴̴k̴ ✞

486,727 просмотров • 10 месяцев назад

Made with Seedance 2.0 Used ~6000 token, 1 person, 2 nights, 6 hours My son and daughter, my dog and a random toy on shelf. Seedance 2.0 is amazing—prompt accuracy is excellent. Scene consistency is trickier, but definitely achievable by anyone. Min Choi PJ Ace Javi Lopez ⛩️ Linus ✦ Ekenstam el.cine Marcio Lima 利真マルシオ 💎 Mark Gadala-Maria AYi Alex Patrascu Amira Zairi 𝐌 LudovicCreator Promise

Made with Seedance 2.0 Used ~6000 token, 1 person, 2 nights, 6 hours My son and daughter, my dog and a random toy on shelf. Seedance 2.0 is amazing—prompt accuracy is excellent. Scene consistency is trickier, but definitely achievable by anyone. Min Choi PJ Ace Javi Lopez ⛩️ Linus ✦ Ekenstam el.cine Marcio Lima 利真マルシオ 💎 Mark Gadala-Maria AYi Alex Patrascu Amira Zairi 𝐌 LudovicCreator Promise

vfxpro88.Ξth

192,191 просмотров • 6 месяцев назад

Mac-1 6.6B searching files on my Macbook with 10/10 accuracy. Mac's Finder and Spotlight are broken and painful to use. I have 650GB of data on my Mac, and it's all scattered. I can't find files manually. Now I ask Mac-1 to find any file (PDF, JSON, PNG, DOC, etc.), and it finds it in 2–4 seconds. Just amazing. To get early access to Mac-1, join the waitlist:

Mac-1 6.6B searching files on my Macbook with 10/10 accuracy. Mac's Finder and Spotlight are broken and painful to use. I have 650GB of data on my Mac, and it's all scattered. I can't find files manually. Now I ask Mac-1 to find any file (PDF, JSON, PNG, DOC, etc.), and it finds it in 2–4 seconds. Just amazing. To get early access to Mac-1, join the waitlist:

CJ Zafir

67,485 просмотров • 3 месяцев назад

The X68000 was one of the best home computers for enjoying classic Konami and Capcom arcade games. Its hardware was so powerful that, aside from the Neo Geo, it was virtually the only system capable of delivering near-perfect arcade conversions at the time. Unfortunately, it was extremely expensive and sold exclusively in Japan, so very few people ever had the chance to experience it. Even today, decades after its release, the quality of its games is still incredibly impressive. Ghouls ’n Ghosts is a perfect example. Compared to the Mega Drive and PC Engine versions, the X68000 port is on an entirely different level in terms of accuracy and overall quality. It’s still amazing to think that a game this good could fit on just two floppy disks.

The X68000 was one of the best home computers for enjoying classic Konami and Capcom arcade games. Its hardware was so powerful that, aside from the Neo Geo, it was virtually the only system capable of delivering near-perfect arcade conversions at the time. Unfortunately, it was extremely expensive and sold exclusively in Japan, so very few people ever had the chance to experience it. Even today, decades after its release, the quality of its games is still incredibly impressive. Ghouls ’n Ghosts is a perfect example. Compared to the Mega Drive and PC Engine versions, the X68000 port is on an entirely different level in terms of accuracy and overall quality. It’s still amazing to think that a game this good could fit on just two floppy disks.

Gamer_lafan

20,452 просмотров • 2 месяцев назад

Israel Dominates the Battle Against Cancer yet Again. Israeli technology is destroying tumors inside the body, without a single cut. Simply amazing. ExAblate by Insightec uses MRI-guided focused ultrasound. Thousands of precise sound waves converge like an acoustic scalpel deep inside the body, heating and destroying cancerous tissue with millimeter accuracy while sparing surrounding healthy areas. It’s already treating locally confined prostate cancer, providing pain relief from bone metastases, and offering new hope for patients who want to avoid surgery or radiation. The same platform is also transforming lives by treating essential tremor and tremor-dominant Parkinson’s without opening the skull. No incisions. No radiation. Real-time MRI precision. This isn’t science fiction anymore. It’s Israeli innovation delivering non-invasive, life-changing treatment for cancer and neurological diseases. Proud every single day. 🇮🇱 #ExAblate #Insightec #FocusedUltrasound #ProstateCancer #IsraelInnovation #StartupNation

Israel Dominates the Battle Against Cancer yet Again. Israeli technology is destroying tumors inside the body, without a single cut. Simply amazing. ExAblate by Insightec uses MRI-guided focused ultrasound. Thousands of precise sound waves converge like an acoustic scalpel deep inside the body, heating and destroying cancerous tissue with millimeter accuracy while sparing surrounding healthy areas. It’s already treating locally confined prostate cancer, providing pain relief from bone metastases, and offering new hope for patients who want to avoid surgery or radiation. The same platform is also transforming lives by treating essential tremor and tremor-dominant Parkinson’s without opening the skull. No incisions. No radiation. Real-time MRI precision. This isn’t science fiction anymore. It’s Israeli innovation delivering non-invasive, life-changing treatment for cancer and neurological diseases. Proud every single day. 🇮🇱 #ExAblate #Insightec #FocusedUltrasound #ProstateCancer #IsraelInnovation #StartupNation

Mor Edge Insight

32,615 просмотров • 4 месяцев назад

🌍Are LLMs aware of cultural and legal safety in today’s geo-diverse world? 🚀Introducing SafeWorld, our #NeurIPS2024 paper and benchmark assessing LLMs’ understanding of geo-diverse safety, based on cultural norms and policies across 50 countries and 493 regions/races. ⚖️We also propose a multi-dimensional framework for evaluating contextual appropriateness, accuracy, and comprehensiveness, revealing major gaps in current LLMs. 🧨To address this, we train SafeWorldLM using DPO, achieving SOTA performance and a 20% higher global human evaluator rating in helpfulness and harmfulness over competing models, including GPT-4o. 🔗Paper: 💻 GitHub: 🫶🏻This is a joint leading effort with Da Yin. Also many thanks to the amazing team Kung-Hsiang Steeve Huang Kai-Wei Chang, and Violet Peng for their hard work. Check out more details and results we conclude from our paper in the thread below. 🧵

🌍Are LLMs aware of cultural and legal safety in today’s geo-diverse world? 🚀Introducing SafeWorld, our #NeurIPS2024 paper and benchmark assessing LLMs’ understanding of geo-diverse safety, based on cultural norms and policies across 50 countries and 493 regions/races. ⚖️We also propose a multi-dimensional framework for evaluating contextual appropriateness, accuracy, and comprehensiveness, revealing major gaps in current LLMs. 🧨To address this, we train SafeWorldLM using DPO, achieving SOTA performance and a 20% higher global human evaluator rating in helpfulness and harmfulness over competing models, including GPT-4o. 🔗Paper: 💻 GitHub: 🫶🏻This is a joint leading effort with Da Yin. Also many thanks to the amazing team Kung-Hsiang Steeve Huang Kai-Wei Chang, and Violet Peng for their hard work. Check out more details and results we conclude from our paper in the thread below. 🧵

Haoyi Qiu

16,344 просмотров • 1 год назад

Two years ago leading up to the draft, Chris Simms had Bo Nix higher than most; Nix was his QB3. So, I followed up with Simms on this to look back at his pre-draft evaluation. "People draft offenses too much. That's the other problem," Simms explained. "So, they draft the offense of the team. That's how the busts happen. That's how Tim Tebow or Matt Leinart happen. Where they go, 'Oh, they win games and their stats are amazing!' And I go, 'Well, watch the game. Tim Tebow threw the ball one yard to Percy Harvin and he ran 80 yards and we go, 'Tim Tebow, amazing!' or Matt Leinart threw a screen to Reggie (Bush) and he ran 100 yards all over the place and we go, 'He's the No. 1 pick; he's a winner!' And that's how you bust on the draft. And with Bo Nix, what happened was, 'Oh, all he does is throw screens.' Well, here we are, Dante Moore, he's still throwing screens, he's got a big arm. That's the offense they run. And there was enough to see and go, 'This is a downfield throw, whoa, the ball popped out of his hand. He can do it with ease.' And then the decision-making and the accuracy. So, I feel like the narrative took over with Bo Nix rather than people just looking at what he is and what he does."

Two years ago leading up to the draft, Chris Simms had Bo Nix higher than most; Nix was his QB3. So, I followed up with Simms on this to look back at his pre-draft evaluation. "People draft offenses too much. That's the other problem," Simms explained. "So, they draft the offense of the team. That's how the busts happen. That's how Tim Tebow or Matt Leinart happen. Where they go, 'Oh, they win games and their stats are amazing!' And I go, 'Well, watch the game. Tim Tebow threw the ball one yard to Percy Harvin and he ran 80 yards and we go, 'Tim Tebow, amazing!' or Matt Leinart threw a screen to Reggie (Bush) and he ran 100 yards all over the place and we go, 'He's the No. 1 pick; he's a winner!' And that's how you bust on the draft. And with Bo Nix, what happened was, 'Oh, all he does is throw screens.' Well, here we are, Dante Moore, he's still throwing screens, he's got a big arm. That's the offense they run. And there was enough to see and go, 'This is a downfield throw, whoa, the ball popped out of his hand. He can do it with ease.' And then the decision-making and the accuracy. So, I feel like the narrative took over with Bo Nix rather than people just looking at what he is and what he does."

Andrew Mason

534,765 просмотров • 7 месяцев назад

This can be the superpower of the Neural Band that Meta is giving together with the Ray-Ban Meta Display glasses. The video shows an old prototype bracelet by CTRL+LABS, the startup acquired by Meta and whose technology was used to develop the Neural Band. At the beginning of the video, the guy makes an action (a keyboard key press) with his hands, then the bracelet can substitute the key pressure, and at the end of the video, the guy doesn't even have to do the action; it is just sufficient that he "thinks" about it. As long as the brain is sending an electric message to the fingers, the full action is not necessary anymore. Just an "intention" to move them is necessary. If the Neural Band is evolved to this stage, and the users are educated to this, potentially, we may not even need to perform air taps or writing gestures, but we could just think about doing them. This would reduce a lot of the fatigue of using XR devices and the weirdness of using them on the street. Then why isn't this feature available today? I guess that the reason is twofold. First of all, we have accuracy: the full gesture is easier to detect for the system. Many people (me included) are praising the accuracy of the Neural Band, and this is amazing, because an input mechanism should have a reliability close to 100%. Then we, as users, have never been trained to just "think" about actions: it would feel weird and hard to learn. I think we should undergo some training to learn how to do this "thinking" operation properly. I hope that something like this could come in the upcoming years... that would be the real game-changer paradigm if compared to the camera-based tracking.

This can be the superpower of the Neural Band that Meta is giving together with the Ray-Ban Meta Display glasses. The video shows an old prototype bracelet by CTRL+LABS, the startup acquired by Meta and whose technology was used to develop the Neural Band. At the beginning of the video, the guy makes an action (a keyboard key press) with his hands, then the bracelet can substitute the key pressure, and at the end of the video, the guy doesn't even have to do the action; it is just sufficient that he "thinks" about it. As long as the brain is sending an electric message to the fingers, the full action is not necessary anymore. Just an "intention" to move them is necessary. If the Neural Band is evolved to this stage, and the users are educated to this, potentially, we may not even need to perform air taps or writing gestures, but we could just think about doing them. This would reduce a lot of the fatigue of using XR devices and the weirdness of using them on the street. Then why isn't this feature available today? I guess that the reason is twofold. First of all, we have accuracy: the full gesture is easier to detect for the system. Many people (me included) are praising the accuracy of the Neural Band, and this is amazing, because an input mechanism should have a reliability close to 100%. Then we, as users, have never been trained to just "think" about actions: it would feel weird and hard to learn. I think we should undergo some training to learn how to do this "thinking" operation properly. I hope that something like this could come in the upcoming years... that would be the real game-changer paradigm if compared to the camera-based tracking.

TonyVT SkarredGhost

11,805 просмотров • 11 месяцев назад

Introducing ✨RigidFormer: Learning Rigid Dynamics with Transformers - our attempt to scale learning-based physical dynamics with Transformers. RigidFormer learns rigid dynamics with Transformers. It is a mesh-free, object-centric Transformer for multi-object rigid-body contact dynamics from point clouds. Learning physics with purely neural simulators, without relying on traditional physics engines, is an important and widely studied problem. Prior SOTA methods often use graph neural networks for accuracy and generalization, but still struggle with efficient, high-fidelity simulation at scale. RigidFormer uses only point inputs, matches or outperforms mesh-based baselines on standard benchmarks, runs much faster, generalizes across point resolutions and datasets, and scales to 200+ objects. We also show a preliminary extension to command-conditioned articulated bodies by treating body parts as interacting object-level components. RigidFormer is mesh-free: it does not require mesh connectivity, SDFs, or vertex-level message passing, making it well-suited for point-cloud observations and scalable simulation. This architecture can also be adapted to learn soft-body dynamics by replacing the rigid-body module (differentiable Kabsch alignment). 🎬See our video for more details. Many thanks to my amazing collaborators: Minghao Guo Minghao Guo, Haixu Wu Haixu Wu 吴海旭, Doug Roble, Tuur Stuyck Tuur Stuyck, and Wojciech Matusik Wojciech Matusik. Project page: Paper:

Introducing ✨RigidFormer: Learning Rigid Dynamics with Transformers - our attempt to scale learning-based physical dynamics with Transformers. RigidFormer learns rigid dynamics with Transformers. It is a mesh-free, object-centric Transformer for multi-object rigid-body contact dynamics from point clouds. Learning physics with purely neural simulators, without relying on traditional physics engines, is an important and widely studied problem. Prior SOTA methods often use graph neural networks for accuracy and generalization, but still struggle with efficient, high-fidelity simulation at scale. RigidFormer uses only point inputs, matches or outperforms mesh-based baselines on standard benchmarks, runs much faster, generalizes across point resolutions and datasets, and scales to 200+ objects. We also show a preliminary extension to command-conditioned articulated bodies by treating body parts as interacting object-level components. RigidFormer is mesh-free: it does not require mesh connectivity, SDFs, or vertex-level message passing, making it well-suited for point-cloud observations and scalable simulation. This architecture can also be adapted to learn soft-body dynamics by replacing the rigid-body module (differentiable Kabsch alignment). 🎬See our video for more details. Many thanks to my amazing collaborators: Minghao Guo Minghao Guo, Haixu Wu Haixu Wu 吴海旭, Doug Roble, Tuur Stuyck Tuur Stuyck, and Wojciech Matusik Wojciech Matusik. Project page: Paper:

Zhiyang (Frank) Dou

573,369 просмотров • 4 месяцев назад

Legora sets the bar for operating at AI speed. Watching them become one of the fastest growing software companies of all time these past few years has provided constant lessons on what’s required to win in this new world. Fresh off Legora's $550M Series D, CEO Max Junestrand joined logan bartlett and me on Unsupervised Learning to provide a masterclass on building an AI-native company. He shared some amazing lessons around - Constantly rebuilding for the bleeding edge of model capabilities - Partnering with customers for both immediate impact and long-term transformation - Running Legora differently from traditional software companies He also included some spicy takes on - Why foundation models entering legal is good for Legora - Pricing AI products - The future of the legal industry It’s impossible to listen to Max and not pick up the infectious energy that makes Legora such a special company. Check out the full episode: YouTube: Spotify: Apple: 0:00 Intro 1:16 Legora’s Series D Story 3:24 Why You Need Low Ego to Build in AI 5:58 From 60% to 100% Accuracy in One Summer 7:04 Law Firm Economics Shift 14:09 Pricing Seats Vs Outcomes 18:31 Why Foundation Models Entering Legal Helps Legora 30:10 Convincing a 75-Year-Old Partner to Go All In 33:02 Hiring Legal Engineers 34:32 Running an AI-Native Company 35:57 The Opus 4.5 Christmas Breakthrough 40:02 Building With Customers 44:01 All In On US Expansion 51:22 Stockholm Startup DNA

Legora sets the bar for operating at AI speed. Watching them become one of the fastest growing software companies of all time these past few years has provided constant lessons on what’s required to win in this new world. Fresh off Legora's $550M Series D, CEO Max Junestrand joined logan bartlett and me on Unsupervised Learning to provide a masterclass on building an AI-native company. He shared some amazing lessons around - Constantly rebuilding for the bleeding edge of model capabilities - Partnering with customers for both immediate impact and long-term transformation - Running Legora differently from traditional software companies He also included some spicy takes on - Why foundation models entering legal is good for Legora - Pricing AI products - The future of the legal industry It’s impossible to listen to Max and not pick up the infectious energy that makes Legora such a special company. Check out the full episode: YouTube: Spotify: Apple: 0:00 Intro 1:16 Legora’s Series D Story 3:24 Why You Need Low Ego to Build in AI 5:58 From 60% to 100% Accuracy in One Summer 7:04 Law Firm Economics Shift 14:09 Pricing Seats Vs Outcomes 18:31 Why Foundation Models Entering Legal Helps Legora 30:10 Convincing a 75-Year-Old Partner to Go All In 33:02 Hiring Legal Engineers 34:32 Running an AI-Native Company 35:57 The Opus 4.5 Christmas Breakthrough 40:02 Building With Customers 44:01 All In On US Expansion 51:22 Stockholm Startup DNA

Jacob Effron

32,227 просмотров • 6 месяцев назад

Opus 4.7 - 400k vs 1m context - is there a difference? I've heard Theo - t3.gg talk about the fact that it is unlikely that Anthropic would have offered up a model with 1m context at the same cost, if it wasn't a different (i.e. cheaper to serve) model. I did a test where I toggled the 1m default model on & off in Claude Code (otherwise default settings, xHigh reasoning) and compared the outputs with 3x generations - same prompts etc. My observations: - Models feel DIFFERENT - often when you ask a model for the same generation, you get a somewhat different answer, but it feels & smells the same. Here 400k and 1m are very different every time - 400k model seems better - not that 1m is trash and 400k is amazing, but there are definitely issues with the level of ambition and accuracy that 1m model seems to have Examples of 1m failing: - Voxel Rome: the colosseum is nowhere near as impressive - Golden Gate: cars go sideways, waves not very high, bridge goes into land; though the structure of the bridge is a bit better - Stonehenge: structure is more 'wrong', lighting, shadows & textures are more flat and not as rich This isn't a conclusive evidence of course, but at least to me the two models do not behave the same way. Anecdotally as well when building 1m felt like it was doing more weird validation (e.g. going around in circles) and 400k was more straightforward. These sorts of things are harder to capture in tests, but you'd notice in Claude Code. You can review the hosted generations, see the code & prompts in the links below

Opus 4.7 - 400k vs 1m context - is there a difference? I've heard Theo - t3.gg talk about the fact that it is unlikely that Anthropic would have offered up a model with 1m context at the same cost, if it wasn't a different (i.e. cheaper to serve) model. I did a test where I toggled the 1m default model on & off in Claude Code (otherwise default settings, xHigh reasoning) and compared the outputs with 3x generations - same prompts etc. My observations: - Models feel DIFFERENT - often when you ask a model for the same generation, you get a somewhat different answer, but it feels & smells the same. Here 400k and 1m are very different every time - 400k model seems better - not that 1m is trash and 400k is amazing, but there are definitely issues with the level of ambition and accuracy that 1m model seems to have Examples of 1m failing: - Voxel Rome: the colosseum is nowhere near as impressive - Golden Gate: cars go sideways, waves not very high, bridge goes into land; though the structure of the bridge is a bit better - Stonehenge: structure is more 'wrong', lighting, shadows & textures are more flat and not as rich This isn't a conclusive evidence of course, but at least to me the two models do not behave the same way. Anecdotally as well when building 1m felt like it was doing more weird validation (e.g. going around in circles) and 400k was more straightforward. These sorts of things are harder to capture in tests, but you'd notice in Claude Code. You can review the hosted generations, see the code & prompts in the links below

Peter Gostev (SF: 22-26 June)

29,203 просмотров • 4 месяцев назад

🚀 We're thrilled to introduce Orthrus 🧬🐕—a groundbreaking mature RNA foundation model designed to push the boundaries of RNA property prediction! 🔬 What is Orthrus? Orthrus is a Mamba-based RNA foundation model, pre-trained using a novel self-supervised contrastive learning objective with biologically inspired augmentations. It optimizes the similarity between splicing isoforms and orthologous transcripts, capturing functional and evolutionary relationships to enhance mature RNA property prediction accuracy. 📑 Preprint: 💻 Code: 🌐 Project Page: 📦 Model Weights: 🧠 Why Orthrus? Decoding the RNA regulatory code is key to understanding biology, but traditional experimental approaches are slow and costly. Existing genomic foundation models rely on techniques like masked language modeling or next-token prediction, which aren't fully aligned with the complexities of genomic data—leading to suboptimal results. 🌟 Orthrus Highlights: - Biologically-Informed Contrastive Learning 🧪: A novel contrastive learning objective designed specifically for genomics, maximizing similarity between splicing isoforms and orthologous transcripts across species. - Extensive Pre-training 📊: Trained on splicing annotations from 10 species and orthologous alignments from 400+ mammalian species (Zoonomia Project), with a focus on sequences of high functional importance. - Superior Representations🏅: Orthrus outperforms existing genomic models on 5 mRNA property prediction tasks, often surpassing supervised methods with just a simple linear transformation. - Efficiency in Low-Data Settings📉: Orthrus excels in low-data regimes, achieving state-of-the-art results with as few as 45 labeled examples for fine-tuning on RNA half-life prediction. Shoutout to the amazing leading authors Phil (Phil Fradkin) and Ian (Ian Shi)! Also the work is impossible without an outstanding collaboration by Karina (Karin(a) Isaev), Brendan (Brendan Frey) , Quaid (Quaid Morris), Leo J. Lee! Vector Institute University Health Network U of T Department of Computer Science Temerty Centre for AI in Medicine (T-CAIREM) Department of Laboratory Medicine & Pathobiology

🚀 We're thrilled to introduce Orthrus 🧬🐕—a groundbreaking mature RNA foundation model designed to push the boundaries of RNA property prediction! 🔬 What is Orthrus? Orthrus is a Mamba-based RNA foundation model, pre-trained using a novel self-supervised contrastive learning objective with biologically inspired augmentations. It optimizes the similarity between splicing isoforms and orthologous transcripts, capturing functional and evolutionary relationships to enhance mature RNA property prediction accuracy. 📑 Preprint: 💻 Code: 🌐 Project Page: 📦 Model Weights: 🧠 Why Orthrus? Decoding the RNA regulatory code is key to understanding biology, but traditional experimental approaches are slow and costly. Existing genomic foundation models rely on techniques like masked language modeling or next-token prediction, which aren't fully aligned with the complexities of genomic data—leading to suboptimal results. 🌟 Orthrus Highlights: - Biologically-Informed Contrastive Learning 🧪: A novel contrastive learning objective designed specifically for genomics, maximizing similarity between splicing isoforms and orthologous transcripts across species. - Extensive Pre-training 📊: Trained on splicing annotations from 10 species and orthologous alignments from 400+ mammalian species (Zoonomia Project), with a focus on sequences of high functional importance. - Superior Representations🏅: Orthrus outperforms existing genomic models on 5 mRNA property prediction tasks, often surpassing supervised methods with just a simple linear transformation. - Efficiency in Low-Data Settings📉: Orthrus excels in low-data regimes, achieving state-of-the-art results with as few as 45 labeled examples for fine-tuning on RNA half-life prediction. Shoutout to the amazing leading authors Phil (Phil Fradkin) and Ian (Ian Shi)! Also the work is impossible without an outstanding collaboration by Karina (Karin(a) Isaev), Brendan (Brendan Frey) , Quaid (Quaid Morris), Leo J. Lee! Vector Institute University Health Network U of T Department of Computer Science Temerty Centre for AI in Medicine (T-CAIREM) Department of Laboratory Medicine & Pathobiology

Bo Wang

114,994 просмотров • 1 год назад