正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

"Generalization means being able to solve problems that the system hasn't been prepared for." Our latest work in #RSS2025 can automatically invent neural networks as state abstractions, which help robots generalize. Check it out here:

Bowen Li

1,194 subscribers

19,717 次观看 • 1 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

5 条评论

Bowen Li 的头像

Bowen Li1 年前

Very grateful for my amazing collaborators and mentors @tomssilver, Dr. Alexander Gray, @smash0190, and @AirLabCMU.

LNK | 宋天佑 | 1.11 的头像

LNK | 宋天佑 | 1.111 年前

Wild behavior

MangoMagic™️ 的头像

MangoMagic™️1 年前

Inventing robots that think on their feet. What could go wrong? #Oops #RoboticPsychics #NeuralNonsense

Mohammed Daba 的头像

Mohammed Daba1 年前

why it would put the platform first in the container table, any logical agent would likely put it in the place they go first then in the other place as it plans as it goes

Lukas Die Kunst 的头像

Lukas Die Kunst1 年前

IVNTR's bilevel approach to neural predicates is interesting. How does its 77% success rate compare to purely symbolic methods on identical tasks?

相关视频

Excited to share our latest work: “Bio-Inspired Plastic Neural Networks for Zero-Shot Out-of-Distribution Generalization in Complex Animal-Inspired Robots” 🪲🦎 We show that Hebbian learning outperforms LSTM-based adaptation for real-world transfer. It even works without domain randomization! It can handle: ✅ Uneven terrain ✅ Morphological damage ✅ Sim-to-real gaps

Excited to share our latest work: “Bio-Inspired Plastic Neural Networks for Zero-Shot Out-of-Distribution Generalization in Complex Animal-Inspired Robots” 🪲🦎 We show that Hebbian learning outperforms LSTM-based adaptation for real-world transfer. It even works without domain randomization! It can handle: ✅ Uneven terrain ✅ Morphological damage ✅ Sim-to-real gaps

Sebastian Risi

14,415 次观看 • 1 年前

Many scientific problems hinge on finding interpretable formulas that fit data, but neural networks are the outright opposite! Check out our recent work that make neural networks modular and interpretable. If you have interesting datasets at hand, we're happy to collaborate!

Many scientific problems hinge on finding interpretable formulas that fit data, but neural networks are the outright opposite! Check out our recent work that make neural networks modular and interpretable. If you have interesting datasets at hand, we're happy to collaborate!

Ziming Liu

62,120 次观看 • 3 年前

“No mortal man, I repeat, can solve the problem of Rivers State. It's not a physical problem. Has the President been able to solve it???” — Pastor David Ibiyeomie says, as he Prays for Peace in Rivers State

“No mortal man, I repeat, can solve the problem of Rivers State. It's not a physical problem. Has the President been able to solve it???” — Pastor David Ibiyeomie says, as he Prays for Peace in Rivers State

Port Harcourt Socials

118,424 次观看 • 9 个月前

"Being Prejudiced Is Awesome" It's the ability handed down by our ancestors to help us survive, but that doesn't mean we should turn off our brains when presented with more relevant information. Your calculations should update when you know more, which is why it is IMPORTANT for people to be able to discuss the data... so we can solve the actual problems for everyone!

"Being Prejudiced Is Awesome" It's the ability handed down by our ancestors to help us survive, but that doesn't mean we should turn off our brains when presented with more relevant information. Your calculations should update when you know more, which is why it is IMPORTANT for people to be able to discuss the data... so we can solve the actual problems for everyone!

Leaflit 🍃 Angel's Sword

18,769 次观看 • 2 个月前

Look at this our FBA brother invented something to help mankind yet Africans only invent problems they can’t even solve openly pooping in the streets no wonder they want us over there to do all the work

Look at this our FBA brother invented something to help mankind yet Africans only invent problems they can’t even solve openly pooping in the streets no wonder they want us over there to do all the work

Chris

179,556 次观看 • 1 年前

Liz Kendall: "... if they repeatedly refuse to take the action required Ofcom can apply for business disruptions, which means they wouldn't be able to have their services available in the UK" Susanna Reid: "And how often as that happened?" "It hasn't happened.."

Liz Kendall: "... if they repeatedly refuse to take the action required Ofcom can apply for business disruptions, which means they wouldn't be able to have their services available in the UK" Susanna Reid: "And how often as that happened?" "It hasn't happened.."

Saul Staniforth

36,545 次观看 • 5 个月前

Elon: "In order to make it work with computers, you need to solve real-world AI and vision. We need cameras and silicon neural nets for self-driving to work in a system designed for eyes and biological neural nets." Elon Musk Writer: Claudio

Elon: "In order to make it work with computers, you need to solve real-world AI and vision. We need cameras and silicon neural nets for self-driving to work in a system designed for eyes and biological neural nets." Elon Musk Writer: Claudio

Mario Nawfal

50,283 次观看 • 3 个月前

Democrat Rep. Al Green: "We have to have reparations. If money can solve problems for everybody else, it can solve some of our problems!"

Democrat Rep. Al Green: "We have to have reparations. If money can solve problems for everybody else, it can solve some of our problems!"

TheBlaze

425,601 次观看 • 5 个月前

Democrat Rep. Al Green: "We have to have reparations. If money can solve problems for everybody else, it can solve some of our problems!"

Democrat Rep. Al Green: "We have to have reparations. If money can solve problems for everybody else, it can solve some of our problems!"

Defiant L’s

37,641 次观看 • 5 个月前

Interested in upgrading your Motion Matching system? Check out our #SIGGRAPH2024 work where we introduce *Codebook Matching* to train neural character controllers for synthesizing lifelike embodied animations in #Games and #VR! Project Page:

Interested in upgrading your Motion Matching system? Check out our #SIGGRAPH2024 work where we introduce Codebook Matching to train neural character controllers for synthesizing lifelike embodied animations in #Games and #VR! Project Page:

Sebastian Starke

18,703 次观看 • 2 年前

We believe AI can be a dedicated research partner to help discover the next breakthrough. Enter Co-Scientist: our latest Gemini-based multi-agent system that can generate, debate and evolve novel hypotheses for complex scientific problems 🧵

We believe AI can be a dedicated research partner to help discover the next breakthrough. Enter Co-Scientist: our latest Gemini-based multi-agent system that can generate, debate and evolve novel hypotheses for complex scientific problems 🧵

Google DeepMind

204,409 次观看 • 5 个月前

Ever wondered what neural networks are and how they work? Systems like ChatGPT use neural networks to work as well as they do. Neural networks are composed of "layers" of neurons, layers with different functions; connections between layers called "weights"; and mathematical functions called "activation functions". If you’re interested in learning about these systems, check the comments. Ultimately, the neural network structure of the model serves to visually demonstrate that it is, in fact, a complex mathematical equation. When companies release the model's weights, they are releasing a key component needed to run the model's complete equation. Without the weights, the equation is incomplete. For the math-minded: the weights of a model are the learned numbers (they are variables during training) that are then used as constants in the mathematical functions that make up the model. Neural networks are ultimately just one big, hyper-complex mathematical function, and when a model is trained, it learns the constants associated with the high-dimensional input.

Ever wondered what neural networks are and how they work? Systems like ChatGPT use neural networks to work as well as they do. Neural networks are composed of "layers" of neurons, layers with different functions; connections between layers called "weights"; and mathematical functions called "activation functions". If you’re interested in learning about these systems, check the comments. Ultimately, the neural network structure of the model serves to visually demonstrate that it is, in fact, a complex mathematical equation. When companies release the model's weights, they are releasing a key component needed to run the model's complete equation. Without the weights, the equation is incomplete. For the math-minded: the weights of a model are the learned numbers (they are variables during training) that are then used as constants in the mathematical functions that make up the model. Neural networks are ultimately just one big, hyper-complex mathematical function, and when a model is trained, it learns the constants associated with the high-dimensional input.

Harper Carroll

50,905 次观看 • 1 年前

People don’t want politicians to pick fights with one another, they want us to work together to solve problems. We’re at the table and want to work with the Provinces to help solve some of the biggest challenges facing the nation, so we can help make Canada the very best country it can be.

People don’t want politicians to pick fights with one another, they want us to work together to solve problems. We’re at the table and want to work with the Provinces to help solve some of the biggest challenges facing the nation, so we can help make Canada the very best country it can be.

Sean Fraser

260,050 次观看 • 2 年前

Drop the camera, and bang -- the robots do the job without camera calibration or additional data collection. We want robots to manipulate anywhere! Our new work, **Maniwhere**, enables robots to manipulate objects from any camera view with any background. Isn't this the generalization every roboticist is looking for? How do we achieve this with sim2real? You can check this thread from Zhecheng Yuan ! And the website is here:

Drop the camera, and bang -- the robots do the job without camera calibration or additional data collection. We want robots to manipulate anywhere! Our new work, Maniwhere, enables robots to manipulate objects from any camera view with any background. Isn't this the generalization every roboticist is looking for? How do we achieve this with sim2real? You can check this thread from Zhecheng Yuan ! And the website is here:

Huazhe Harry Xu

28,935 次观看 • 2 年前

As we're celebrating our 16th birthday, we're reminded that our team is empowered to work together, solve hard problems, and act in the best interest of users to help build a better Internet. What is your favorite tech company or technology, and why do you love it? #BirthdayWeek

As we're celebrating our 16th birthday, we're reminded that our team is empowered to work together, solve hard problems, and act in the best interest of users to help build a better Internet. What is your favorite tech company or technology, and why do you love it? #BirthdayWeek

Cloudflare

11,576 次观看 • 1 个月前