Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Kalman Filtering + Deep Attention Models = LOVE ❤️ #CoRL '23 paper: we identify a link between Kalman filters and neural attention mechanisms. Idea: the Kalman-gain is a form of attention! No need to tweak parameters. Check website and code: ASU Robotics

asurobot's profile picture

Heni Ben Amor

1,464 subscribers

19,070 Aufrufe • vor 2 Jahren •via X (Twitter)

Wissenschaft & TechnologieBildungGesundheit & Wellness#CoRL
LIVE

Anya Rossi• Live Now

Private livecam show

2 Kommentare

Profilbild von Oliver Obst
Oliver Obstvor 2 Jahren

@ASURobotics Hey you guys are too quick :) congratulations

Profilbild von Z❤️H
Z❤️Hvor 2 Jahren

@ASURobotics Every hidden state should be a latent space

Ähnliche Videos

New short course: Attention in Transformers: Concepts and Code in PyTorch. Last week we released a course on how LLM transformers work. This week, go deeper and learn about the technical ideas behind the attention mechanism, and see how to code it in PyTorch. This course is built with Joshua Starmer, Founder and CEO of StatQuest. The attention mechanism was a breakthrough that led to transformers, the architecture powering large language models like ChatGPT. Transformers, introduced in the 2017 paper: "Attention is All You Need" by Viswani and others, took off because of its highly scalable design. In this course, you’ll learn how the attention mechanism, a key element of transformer-based LLMs, works and implement it in PyTorch. You'll develop deep intuition about building reliable, functional, and scalable AI applications. What you will do: - Understand the evolution of the attention mechanism, a key breakthrough that led to transformers. - Learn the relationships between word embeddings, positional embeddings, and attention. - Learn about the Query, Key, and Value matrices, and how to produce and use them in attention. - Walk through the math required to calculate self-attention and masked self-attention to learn why and how they work. - Understand the difference between self-attention and masked self-attention and how one is used in the encoder to build context-aware embeddings and the other is used in the decoder for generative outputs. - Learn the details of the encoder-decoder architecture, cross-attention, and multi-head attention and how they are all incorporated into a transformer. - Use PyTorch to code a class that implements self-attention, masked self-attention, and multi-head attention. There're lots of exciting technical details in this course. Please sign up here:
2:41

New short course: Attention in Transformers: Concepts and Code in PyTorch. Last week we released a course on how LLM transformers work. This week, go deeper and learn about the technical ideas behind the attention mechanism, and see how to code it in PyTorch. This course is built with Joshua Starmer, Founder and CEO of StatQuest. The attention mechanism was a breakthrough that led to transformers, the architecture powering large language models like ChatGPT. Transformers, introduced in the 2017 paper: "Attention is All You Need" by Viswani and others, took off because of its highly scalable design. In this course, you’ll learn how the attention mechanism, a key element of transformer-based LLMs, works and implement it in PyTorch. You'll develop deep intuition about building reliable, functional, and scalable AI applications. What you will do: - Understand the evolution of the attention mechanism, a key breakthrough that led to transformers. - Learn the relationships between word embeddings, positional embeddings, and attention. - Learn about the Query, Key, and Value matrices, and how to produce and use them in attention. - Walk through the math required to calculate self-attention and masked self-attention to learn why and how they work. - Understand the difference between self-attention and masked self-attention and how one is used in the encoder to build context-aware embeddings and the other is used in the decoder for generative outputs. - Learn the details of the encoder-decoder architecture, cross-attention, and multi-head attention and how they are all incorporated into a transformer. - Use PyTorch to code a class that implements self-attention, masked self-attention, and multi-head attention. There're lots of exciting technical details in this course. Please sign up here:

Andrew Ng

132,400 Aufrufe • vor 1 Jahr

The code of GSPN #CVPR2025 is released! We proposed a new sqrt(N) complexity attention mechanism, which enables efficient high resolution image generation. We can generate 8k images with 42x speed up compared to self-attention in StableDiffusionXL! Code: Paper:
0:35

The code of GSPN #CVPR2025 is released! We proposed a new sqrt(N) complexity attention mechanism, which enables efficient high resolution image generation. We can generate 8k images with 42x speed up compared to self-attention in StableDiffusionXL! Code: Paper:

Xiaolong Wang

354,887 Aufrufe • vor 1 Jahr

When did police start dressing like Fortnite characters? And what is camouflage good for in a city? Dress is a form of speech and we need to pay attention to the message we’re being sent.
2:27

When did police start dressing like Fortnite characters? And what is camouflage good for in a city? Dress is a form of speech and we need to pay attention to the message we’re being sent.

Kyle Whitmire

245,468 Aufrufe • vor 7 Monaten

Zhilin at GTC: Introducing Attention Residuals Learning selective memory, rather than mechanically accumulating everything, is the beauty of attention. Many of you have probably read Attention Is All You Need, the 2017 Transformer paper that brought “human-like” attention into the model’s field of view. From that point on, models no longer simply read everything in a mechanical way. Instead, they began to develop a sense of what matters more and what matters less across the text, choosing to retain the more important information. Recently, Kimi applied this idea of attention to the temporal dimension, then rotated it 90 degrees into the model’s depth dimension. This allows the model to have attention not only over time, but also throughout the process of information transmission across layers—giving it a more intelligent way to understand and process information.
5:36

Zhilin at GTC: Introducing Attention Residuals Learning selective memory, rather than mechanically accumulating everything, is the beauty of attention. Many of you have probably read Attention Is All You Need, the 2017 Transformer paper that brought “human-like” attention into the model’s field of view. From that point on, models no longer simply read everything in a mechanical way. Instead, they began to develop a sense of what matters more and what matters less across the text, choosing to retain the more important information. Recently, Kimi applied this idea of attention to the temporal dimension, then rotated it 90 degrees into the model’s depth dimension. This allows the model to have attention not only over time, but also throughout the process of information transmission across layers—giving it a more intelligent way to understand and process information.

Kimi.ai

115,162 Aufrufe • vor 5 Monaten

🇺🇸 Chamath just revealed the truth on Rogan… attention is what runs it all. Google ranks the web by attention. Facebook and Instagram are built to capture attention. The most important AI paper ever written is literally called "Attention Is All You Need." Every system that now runs the world reduces to the same single currency. And at some point, that stops sounding like branding and starts sounding like a warning. Joe Rogan, Chamath Palihapitiya
1:56

🇺🇸 Chamath just revealed the truth on Rogan… attention is what runs it all. Google ranks the web by attention. Facebook and Instagram are built to capture attention. The most important AI paper ever written is literally called "Attention Is All You Need." Every system that now runs the world reduces to the same single currency. And at some point, that stops sounding like branding and starts sounding like a warning. Joe Rogan, Chamath Palihapitiya

Mario Nawfal

314,947 Aufrufe • vor 3 Monaten

Attention, Commander! Code Red situation! Dynasty Forces are mounting a substantial assault on our perimeter. This is a major attack, and we need to be on high alert. #TempestRising
1:38

Attention, Commander! Code Red situation! Dynasty Forces are mounting a substantial assault on our perimeter. This is a major attack, and we need to be on high alert. #TempestRising

Tempest Rising

105,198 Aufrufe • vor 3 Jahren

No way you can tell me this is not a global banger. Raver BNX is a top talent we should pay attention to ❤️🔥🔥 Song title : Hmm Click on the link and enjoy :
1:00

No way you can tell me this is not a global banger. Raver BNX is a top talent we should pay attention to ❤️🔥🔥 Song title : Hmm Click on the link and enjoy :

Unruly King 👑

25,170 Aufrufe • vor 6 Monaten

Google's "Attention is All You Need" paper came from trying to get a 3% gain in Google Translate. Innovation is a consequence of production. "If you don't make the thing, you cede your opportunity to innovate on the thing." ~ Palantir's CTO Shyam Sankar
1:47

Google's "Attention is All You Need" paper came from trying to get a 3% gain in Google Translate. Innovation is a consequence of production. "If you don't make the thing, you cede your opportunity to innovate on the thing." ~ Palantir's CTO Shyam Sankar

Rohan Paul

221,104 Aufrufe • vor 4 Monaten

(1/5) FP4 hardware is here, but 4-bit attention still kills model quality, blocking true end-to-end FP4 serving. To fix that, we propose Attn-QAT, the first systematic study of quantization-aware training for attention. The result: FP4 attention quality is comparable to BF16 attention with 1.1x–1.5x higher throughput than SageAttention3 on an RTX 5090 and 1.39x speedup over FlashAttention-4 on a B200. Blog: Code: Checkpoints:
0:25

(1/5) FP4 hardware is here, but 4-bit attention still kills model quality, blocking true end-to-end FP4 serving. To fix that, we propose Attn-QAT, the first systematic study of quantization-aware training for attention. The result: FP4 attention quality is comparable to BF16 attention with 1.1x–1.5x higher throughput than SageAttention3 on an RTX 5090 and 1.39x speedup over FlashAttention-4 on a B200. Blog: Code: Checkpoints:

Hao AI Lab

37,869 Aufrufe • vor 4 Monaten

Just read a neat AI paper called SAMURAI -- it takes SAM 2 (Meta's "segment anything" model) and makes it way better at tracking objects in videos. Basic problem is SAM 2 gets confused when things move fast or there's a crowd of similar objects (big problem for VFX and video intelligence alike). The fix? They basically gave it a sense of motion and smarter memory selection using our old friend the Kalman filter -- helps it remember where things were going instead of just where they are. No retraining needed and runs in realtime. Elegant solution that matches (and even beats!) a lot of purpose built tracking systems. Nice example of how sometimes you don't need fancy new architectures -- just smart ways to use what the model already knows, plus some tried & tested classical methods. Every time someone's like "we solved tracking!" I check the methods section and... yep, there's our boy Kalman, still crushing it after all these years. Sometimes the old school stuff just works.
0:27

Just read a neat AI paper called SAMURAI -- it takes SAM 2 (Meta's "segment anything" model) and makes it way better at tracking objects in videos. Basic problem is SAM 2 gets confused when things move fast or there's a crowd of similar objects (big problem for VFX and video intelligence alike). The fix? They basically gave it a sense of motion and smarter memory selection using our old friend the Kalman filter -- helps it remember where things were going instead of just where they are. No retraining needed and runs in realtime. Elegant solution that matches (and even beats!) a lot of purpose built tracking systems. Nice example of how sometimes you don't need fancy new architectures -- just smart ways to use what the model already knows, plus some tried & tested classical methods. Every time someone's like "we solved tracking!" I check the methods section and... yep, there's our boy Kalman, still crushing it after all these years. Sometimes the old school stuff just works.

Bilawal Sidhu

47,246 Aufrufe • vor 1 Jahr

Web3 feels empty. It’s not the tech. It’s the people, and we need a lot of them. That’s why we raised $17M to fix the attention and distribution layer. (1/🧵)
0:44

Web3 feels empty. It’s not the tech. It’s the people, and we need a lot of them. That’s why we raised $17M to fix the attention and distribution layer. (1/🧵)

Rexy

470,681 Aufrufe • vor 1 Jahr

one of the most dangerous manipulation tactics is when someone hurts themselves to control others.. It’s narcissism disguised as victimhood, pretending to be a victim to gain attention and sympathy...
1:09

one of the most dangerous manipulation tactics is when someone hurts themselves to control others.. It’s narcissism disguised as victimhood, pretending to be a victim to gain attention and sympathy...

Tweet Chor👑

44,120 Aufrufe • vor 5 Monaten

I love Jeremy's idea of the billion dollar PDF. Every so often someone crystallizes an idea at just the right moment. It sets the narrative for that era and billions of dollars organize around it. - The Bitcoin white paper - Situational Awareness - Attention Is All You Need - Brian Arthur's Increasing Returns - Software is Eating the World
1:02

I love Jeremy's idea of the billion dollar PDF. Every so often someone crystallizes an idea at just the right moment. It sets the narrative for that era and billions of dollars organize around it. - The Bitcoin white paper - Situational Awareness - Attention Is All You Need - Brian Arthur's Increasing Returns - Software is Eating the World

Patrick OShaughnessy

84,673 Aufrufe • vor 1 Monat

Ninamori_crafts brought the 7 Days to Die workbench to life with a handcrafted miniature. The attention to detail is amazing! Show her some love and check out her amazing work
1:48

Ninamori_crafts brought the 7 Days to Die workbench to life with a handcrafted miniature. The attention to detail is amazing! Show her some love and check out her amazing work

7 Days to Die Official

44,429 Aufrufe • vor 1 Monat

I’m convinced the ‘pull over’ scene in Bridesmaids is the most accurate depiction of a woman’s brain during a mid-life crisis. No logic, just vibes and a desperate need for attention. 🎬🎥 Bridesmaids ❤️🔥
2:07

I’m convinced the ‘pull over’ scene in Bridesmaids is the most accurate depiction of a woman’s brain during a mid-life crisis. No logic, just vibes and a desperate need for attention. 🎬🎥 Bridesmaids ❤️🔥

BlueBriefing🎬🍿

29,691 Aufrufe • vor 3 Monaten

End-to-end neural networks racing drones in Abu Dhabi! 🚁 Check out the drone racing team from Delft University of Technology! A completely end-to-end neural network solution, from pixels to direct motor commands. No Kalman filters. No computer vision feature detectors. Just neurons flying the drone. The challenge is extreme. These drones fly at high speeds and need split-second decisions with minimal onboard resources: a single rolling-shutter camera and an IMU. Their approach is called SkyDreamer, based on the Dreamer-v3 reinforcement learning algorithm. First, a world model is trained in simulation. Then, the neural network learns how to fly in its dreams through reinforcement learning. The network's internal state can be read out to see where it thinks it is on the track or how fast it's going. Even better, the drone estimates some of its own body characteristics during flight, like the camera angle relative to the body, eliminating time-consuming manual calibration. The system uses only a single camera and the gyros from the IMU, ignoring the accelerometers, just like human FPV pilots do. ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →
1:01

End-to-end neural networks racing drones in Abu Dhabi! 🚁 Check out the drone racing team from Delft University of Technology! A completely end-to-end neural network solution, from pixels to direct motor commands. No Kalman filters. No computer vision feature detectors. Just neurons flying the drone. The challenge is extreme. These drones fly at high speeds and need split-second decisions with minimal onboard resources: a single rolling-shutter camera and an IMU. Their approach is called SkyDreamer, based on the Dreamer-v3 reinforcement learning algorithm. First, a world model is trained in simulation. Then, the neural network learns how to fly in its dreams through reinforcement learning. The network's internal state can be read out to see where it thinks it is on the track or how fast it's going. Even better, the drone estimates some of its own body characteristics during flight, like the camera angle relative to the body, eliminating time-consuming manual calibration. The system uses only a single camera and the gyros from the IMU, ignoring the accelerometers, just like human FPV pilots do. ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

207,500 Aufrufe • vor 6 Monaten

Enough of calling yourself or someone you love ❤️ a loser - you’re not a loser … you’re just losing … pay attention - listen carefully … watch to the end and send this to someone u love
0:34

Enough of calling yourself or someone you love ❤️ a loser - you’re not a loser … you’re just losing … pay attention - listen carefully … watch to the end and send this to someone u love

Gary Vaynerchuk

34,678 Aufrufe • vor 10 Monaten

Listen: we have to take Trump seriously because, unfortunately, he is the actual President. Trump is a joke, and not always a funny one. But we need to call attention to his rambling and incoherent rants. This was ONE press conference.
1:31

Listen: we have to take Trump seriously because, unfortunately, he is the actual President. Trump is a joke, and not always a funny one. But we need to call attention to his rambling and incoherent rants. This was ONE press conference.

Senator Patty Murray

479,844 Aufrufe • vor 1 Jahr