Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Congratulations also to Patrick 👍 for his #ICLR paper on Temporal Difference (TD) learning , in it We solve the decades-old puzzle of why TD can solve complex RL tasks that Gradient Descent cannot.

thuereyGroup's profile picture

Thuerey Group at TUM

2,817 subscribers

22,074 Aufrufe • vor 1 Jahr •via X (Twitter)

Wissenschaft & TechnologieBildungGesundheit & Wellness#ICLR
LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

RL is back! But is it always the best choice? In a new paper, we investigate under what circumstances neuroevolution outperforms RL in transfer learning tasks. See more details in the thread below 🧵 While NE performs best in simpler domains, it will be interesting to see if the lessons learned here can also be applied to more complex systems/tasks (LLMs?).
0:13

RL is back! But is it always the best choice? In a new paper, we investigate under what circumstances neuroevolution outperforms RL in transfer learning tasks. See more details in the thread below 🧵 While NE performs best in simpler domains, it will be interesting to see if the lessons learned here can also be applied to more complex systems/tasks (LLMs?).

Sebastian Risi

11,130 Aufrufe • vor 11 Monaten

Tick Tock ⏰ Can you solve the clock puzzle?
0:16

Tick Tock ⏰ Can you solve the clock puzzle?

Esha

74,407 Aufrufe • vor 5 Monaten

Can Indy solve the Serpent Puzzle? The Order of Giants DLC is available now for #IndianaJones and the Great Circle.
0:13

Can Indy solve the Serpent Puzzle? The Order of Giants DLC is available now for #IndianaJones and the Great Circle.

machinegames

45,700 Aufrufe • vor 9 Monaten

1/ While most RL methods use shallow MLPs (~2–5 layers), we show that scaling up to 1000-layers for contrastive RL (CRL) can significantly boost performance, ranging from doubling performance to 50x on a diverse suite of robotic tasks. Webpage+Paper+Code:
0:15

1/ While most RL methods use shallow MLPs (~2–5 layers), we show that scaling up to 1000-layers for contrastive RL (CRL) can significantly boost performance, ranging from doubling performance to 50x on a diverse suite of robotic tasks. Webpage+Paper+Code:

Kevin Wang

154,868 Aufrufe • vor 1 Jahr

Can Vision-Language Models Solve the Shell Game? paper:
0:12

Can Vision-Language Models Solve the Shell Game? paper:

AK

38,459 Aufrufe • vor 3 Monaten

Congrats to Mario 👏 for the accept of his ICLR paper on "Diffusion Graph Nets", predicting complex distributions of flow states on unstructured meshes Highly efficient, works even if the training data has only a fraction of the flow statistics.
0:14

Congrats to Mario 👏 for the accept of his ICLR paper on "Diffusion Graph Nets", predicting complex distributions of flow states on unstructured meshes Highly efficient, works even if the training data has only a fraction of the flow statistics.

Thuerey Group at TUM

24,996 Aufrufe • vor 1 Jahr

Congratulations to QB Grant Jordan, earning the College GameDay helmet sticker for his 7 TD performance on Saturday! #Team151
0:14

Congratulations to QB Grant Jordan, earning the College GameDay helmet sticker for his 7 TD performance on Saturday! #Team151

Yale Football

19,137 Aufrufe • vor 1 Jahr

WE WON $5000 SOLVING THE RUST PUZZLE IN theburntpeanut SERVER!💰 GG's to everyone! WAS SO MUCH FUN TO SOLVE!!🎯
0:17

WE WON $5000 SOLVING THE RUST PUZZLE IN theburntpeanut SERVER!💰 GG's to everyone! WAS SO MUCH FUN TO SOLVE!!🎯

Autumn

35,472 Aufrufe • vor 2 Monaten

Jonathan Taylor runs it in for his 2nd TD of the day! LVvsIND on FOX/FOX One
0:11

Jonathan Taylor runs it in for his 2nd TD of the day! LVvsIND on FOX/FOX One

NFL

337,980 Aufrufe • vor 9 Monaten

Robots struggle with strict action rules…memory and symbols help them learn fast. [Project + Full video link ⬇️] Robots struggle when tasks require specific steps in a fixed order. What if memory helped them think symbolically and learn faster? Solving tasks like unlocking a door then opening it is hard for deep RL. But by learning constraint relationships and storing them in memory, robots can solve these tasks much faster; with fewer trials and less training. Why it works ✅ Learns symbolic rules about action constraints ✅ Uses memory to transfer what it learned across tasks ✅ Handles real-world exploration with just 30 minutes of data ✅ Needs 10x fewer episodes than deep RL approaches This memory-based method shows a promising path forward for robots learning structured, real-world tasks. Full video: Paper: Thank you, Mrinal Verghese for sharing this amazing work! 🙏
0:18

Robots struggle with strict action rules…memory and symbols help them learn fast. [Project + Full video link ⬇️] Robots struggle when tasks require specific steps in a fixed order. What if memory helped them think symbolically and learn faster? Solving tasks like unlocking a door then opening it is hard for deep RL. But by learning constraint relationships and storing them in memory, robots can solve these tasks much faster; with fewer trials and less training. Why it works ✅ Learns symbolic rules about action constraints ✅ Uses memory to transfer what it learned across tasks ✅ Handles real-world exploration with just 30 minutes of data ✅ Needs 10x fewer episodes than deep RL approaches This memory-based method shows a promising path forward for robots learning structured, real-world tasks. Full video: Paper: Thank you, Mrinal Verghese for sharing this amazing work! 🙏

Ilir Aliu - eu/acc

10,241 Aufrufe • vor 1 Jahr

My favorite paper this year: "Video models are zero-shot learners and reasoners" It illustrates that video models show emergent visual reasoning at scale - they can solve vision tasks they weren't trained for. This may be the "GPT moment" for vision. Let's break it down 👇
0:19

My favorite paper this year: "Video models are zero-shot learners and reasoners" It illustrates that video models show emergent visual reasoning at scale - they can solve vision tasks they weren't trained for. This may be the "GPT moment" for vision. Let's break it down 👇

Justine Moore

185,001 Aufrufe • vor 6 Monaten

It’s like trying to solve a puzzle cube .
0:13

It’s like trying to solve a puzzle cube .

Valkyrie Brynn

82,933 Aufrufe • vor 1 Jahr

Be honest… How long would it take you to solve this puzzle? 🤔🧩
0:15

Be honest… How long would it take you to solve this puzzle? 🤔🧩

Enezator

25,409 Aufrufe • vor 7 Monaten

Patrick Mahomes to Jason Brownlee for a beautiful back shoulder TD! Stream on @NFLPLus
0:14

Patrick Mahomes to Jason Brownlee for a beautiful back shoulder TD! Stream on @NFLPLus

NFL

879,592 Aufrufe • vor 11 Monaten

Kenneth Gainwell gets his second TD of the day MINvsPIT on NFL Network Also streaming on NFL+
0:11

Kenneth Gainwell gets his second TD of the day MINvsPIT on NFL Network Also streaming on NFL+

NFL

252,910 Aufrufe • vor 9 Monaten

Patrick Mahomes finds Noah Gray for the TD 📺: #KCvsCLE on CBS/Paramount+ 📱:
0:13

Patrick Mahomes finds Noah Gray for the TD 📺: #KCvsCLE on CBS/Paramount+ 📱:

NFL

405,182 Aufrufe • vor 1 Jahr

Gradient experiment in Paper and I genuinely cannot stop
0:14

Gradient experiment in Paper and I genuinely cannot stop

Praveen Kumar

50,984 Aufrufe • vor 3 Monaten

How can we solve the issue with trump? Most of yall:
0:11

How can we solve the issue with trump? Most of yall:

🤍𝕬𝖑𝖎𝖈𝖊🤍

2,505,872 Aufrufe • vor 1 Jahr

If can Tesla AI can solve the most infamous round about in the world in France, they can solve India.
0:13

If can Tesla AI can solve the most infamous round about in the world in France, they can solve India.

Nic Cruz Patane

144,415 Aufrufe • vor 1 Jahr

You CANT solve Suzy Snacktime’s puzzle 🍭
0:15

You CANT solve Suzy Snacktime’s puzzle 🍭

Kandyland

260,993 Aufrufe • vor 1 Jahr