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。

🚨 The biggest challenge for interactive world models isn't image quality. It's drift Small prediction errors compound over time until the generated world falls apart. That's exactly what LingBot-World 2.0 (LingBot-World-Infinity) sets out to solve. A few things that stood out to me 🧵👇

Md Riyazuddin

112,621 subscribers

51,174 次观看 • 7 天前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

One of the biggest challenges LingBot-World 2.0 (Infinity) is trying to address isn't image quality. It's keeping an interactive world coherent over long periods. Once a model starts conditioning on its own predictions, small errors accumulate, eventually leading to visual drift. Here's how it works:

One of the biggest challenges LingBot-World 2.0 (Infinity) is trying to address isn't image quality. It's keeping an interactive world coherent over long periods. Once a model starts conditioning on its own predictions, small errors accumulate, eventually leading to visual drift. Here's how it works:

Markandey Sharma

72,478 次观看 • 8 天前

LingBot-World 2.0 (Infinity) is out on Hugging Face interactive world model with: Hour-long generation with zero quality drift Rich actions & events: attack, cast spells, shoot, summon storms Agentic world: a Director Agent drives real-time world evolution 720p/60fps. Playable like a game

LingBot-World 2.0 (Infinity) is out on Hugging Face interactive world model with: Hour-long generation with zero quality drift Rich actions & events: attack, cast spells, shoot, summon storms Agentic world: a Director Agent drives real-time world evolution 720p/60fps. Playable like a game

AK

28,721 次观看 • 9 天前

Most AI world models impress you for 20 seconds. The harder problem is making them stay coherent. That's what stood out to me in Robbyant's LingBot-World-Infinity. Instead of chasing another resolution headline, the paper focuses on reducing autoregressive drift the issue where small prediction errors compound until the world slowly falls apart. Their stress test is a single uninterrupted 60-minute rollout across 20 different scenarios with no visible quality decay. If you want to see the interactive system yourself, it's live on Reactor: 🎮 I also like that they don't pretend it's perfect. The paper openly says it doesn't have true long-term memory. Leave an area and come back later, and it regenerates something plausible instead of recalling the exact same place. That's a refreshingly honest limitation. Open weights make it even more interesting to dig into. Robbyant For anyone who wants to explore the research: 💻 GitHub: 🤗 Model Weights:

Most AI world models impress you for 20 seconds. The harder problem is making them stay coherent. That's what stood out to me in Robbyant's LingBot-World-Infinity. Instead of chasing another resolution headline, the paper focuses on reducing autoregressive drift the issue where small prediction errors compound until the world slowly falls apart. Their stress test is a single uninterrupted 60-minute rollout across 20 different scenarios with no visible quality decay. If you want to see the interactive system yourself, it's live on Reactor: 🎮 I also like that they don't pretend it's perfect. The paper openly says it doesn't have true long-term memory. Leave an area and come back later, and it regenerates something plausible instead of recalling the exact same place. That's a refreshingly honest limitation. Open weights make it even more interesting to dig into. Robbyant For anyone who wants to explore the research: 💻 GitHub: 🤗 Model Weights:

Shruti Codes

15,066 次观看 • 7 天前

I've been reading through the LingBot-World 2.0 paper today, and I think one detail deserves more attention than the hour-long demo itself. Most interactive world models can generate impressive moments, but keeping an explorable world coherent over long sessions is a much harder problem errors gradually accumulate. LingBot-World 2.0 tackles that challenge by upgrading both world prediction and interaction, sustaining hour-level generation at stable 720p and 60 fps while remaining responsive in real time. The paper argues this comes from its causal training strategy, rather than simply extending generation length. Robbyant

I've been reading through the LingBot-World 2.0 paper today, and I think one detail deserves more attention than the hour-long demo itself. Most interactive world models can generate impressive moments, but keeping an explorable world coherent over long sessions is a much harder problem errors gradually accumulate. LingBot-World 2.0 tackles that challenge by upgrading both world prediction and interaction, sustaining hour-level generation at stable 720p and 60 fps while remaining responsive in real time. The paper argues this comes from its causal training strategy, rather than simply extending generation length. Robbyant

FELIX

113,396 次观看 • 6 天前

Robbyant just open-sourced LingBot-VLA 2.0 and it's much more than a bigger Vision-Language-Action model. Behind the 60,000-hour pretraining dataset run are several clever engineering ideas that push robots closer to reliable real-world deployment. Here are 6 things that stood out to me 👇

Robbyant just open-sourced LingBot-VLA 2.0 and it's much more than a bigger Vision-Language-Action model. Behind the 60,000-hour pretraining dataset run are several clever engineering ideas that push robots closer to reliable real-world deployment. Here are 6 things that stood out to me 👇

Md Riyazuddin

54,251 次观看 • 8 天前

Most world models fall apart after a few seconds. Common failure modes include texture smearing, warped geometry, and scenes that no longer look real. LingBot-World 2.0 from Robbyant seems to hold 720p at 60 fps for a full hour of interaction. That’s impressive. Here is what makes that possible.

Most world models fall apart after a few seconds. Common failure modes include texture smearing, warped geometry, and scenes that no longer look real. LingBot-World 2.0 from Robbyant seems to hold 720p at 60 fps for a full hour of interaction. That’s impressive. Here is what makes that possible.

elvis

11,920 次观看 • 3 天前

🚨🇨🇳HOW CHINA WENT FROM ONE OF THE POOREST COUNTRIES IN THE WORLD TO THE BIGGEST ECONOMY IN THE WORLD (Adjusted for PPP), AND LIFTED OVER 850 MILLION PEOPLE OUT OF POVERTY! ✅WHAT HAS CHINA DONE FOR LESS FORTUNE COUNTRIES? #China #US #Multipolar #World

🚨🇨🇳HOW CHINA WENT FROM ONE OF THE POOREST COUNTRIES IN THE WORLD TO THE BIGGEST ECONOMY IN THE WORLD (Adjusted for PPP), AND LIFTED OVER 850 MILLION PEOPLE OUT OF POVERTY! ✅WHAT HAS CHINA DONE FOR LESS FORTUNE COUNTRIES? #China #US #Multipolar #World

Truth_teller 🇷🇺

15,438 次观看 • 2 年前

More jobs, more food, more hope: that's exactly what World Food Programme resilience programs are delivering around the world. Check out how World Food Programme_WAfrica is working with local communities to literally plant the seeds of a brighter future for families across the #Sahel!

More jobs, more food, more hope: that's exactly what World Food Programme resilience programs are delivering around the world. Check out how World Food Programme_WAfrica is working with local communities to literally plant the seeds of a brighter future for families across the #Sahel!

Carl Skau

13,402 次观看 • 3 年前

Labor has presided over the biggest collapse in living standards in the developed world. It's time the Prime Minister set out a plan to fix this mess.

Labor has presided over the biggest collapse in living standards in the developed world. It's time the Prime Minister set out a plan to fix this mess.

Sussan Ley

66,438 次观看 • 10 个月前

🤖Robots that think ahead and act in real time. LingBot-VA 2.0 — the first embodied-native foundation model. Not fine-tuned from a video generator. Built from scratch for the physical world. ✅ 93.6% success on bimanual tasks ⚡ 150 Hz single-GPU inference 🎯 20 demos to generalize. This is what happens when you stop adapting and start building natively. 🧵👇

🤖Robots that think ahead and act in real time. LingBot-VA 2.0 — the first embodied-native foundation model. Not fine-tuned from a video generator. Built from scratch for the physical world. ✅ 93.6% success on bimanual tasks ⚡ 150 Hz single-GPU inference 🎯 20 demos to generalize. This is what happens when you stop adapting and start building natively. 🧵👇

Robbyant

877,173 次观看 • 8 天前

🚨 Albums Are Prayed Over DEMONICALLY Before Being Sent Out to the World, Claims Boyzone Singer Shane Lynch "I have been in rooms at the top of the top, albums are prayed over demonically that go out to the world" "It's a SATANIC Music Industry"

Sensitive content

🚨 Albums Are Prayed Over DEMONICALLY Before Being Sent Out to the World, Claims Boyzone Singer Shane Lynch "I have been in rooms at the top of the top, albums are prayed over demonically that go out to the world" "It's a SATANIC Music Industry"

Jay Anderson

189,595 次观看 • 1 年前

1/ Tencent releases HY World 1.5 (WorldPlay), an open-source real-time world model with 24fps A streaming video diffusion model that enables real-time, interactive world modeling with long-term geometric consistency In direct comparison to Genie 3, it's technically inferior, and there's no information on how long the consistency will last. However, Tencent has released a significant world model as open source for the first time, making it available to everyone. A few more examples, a thread 🧵

1/ Tencent releases HY World 1.5 (WorldPlay), an open-source real-time world model with 24fps A streaming video diffusion model that enables real-time, interactive world modeling with long-term geometric consistency In direct comparison to Genie 3, it's technically inferior, and there's no information on how long the consistency will last. However, Tencent has released a significant world model as open source for the first time, making it available to everyone. A few more examples, a thread 🧵

Chubby♨️

37,776 次观看 • 7 个月前

***OUT TODAY*** After three extraordinary years following a thread that led from Conservative HQ to Putin's St Petersburg via Kathmandu and a royal Scottish retreat, my new book CUCKOOLAND is out. It's the tale of a world where the rich own the truth. Our world. 👇🧵

OUT TODAY After three extraordinary years following a thread that led from Conservative HQ to Putin's St Petersburg via Kathmandu and a royal Scottish retreat, my new book CUCKOOLAND is out. It's the tale of a world where the rich own the truth. Our world. 👇🧵

Tom Burgis

423,018 次观看 • 2 年前

The robot’s “eyes” just received a big upgrade. LingBot-Depth 2.0, a depth-completion model with half the depth error just dropped. 12/16 benchmarks topped. Glass, mirrors, and transparent objects are so easy for us humans, but so hard for robots, because they do not behave like ordinary surfaces in a camera pipeline. A robot that misunderstands a balcony window or a table edge, will have a completely false planning inside a false world. Huge implecation. LingBot-Depth 2.0 takes an RGB image plus a broken depth map from a sensor and then outputs a cleaner depth map and a usable 3D point cloud. Numbers on LingBot-Depth 2.0 - Excels on glass, mirrors & transparent objects — where traditional depth cameras fail - Training data: 3M → 150M (50x scale-up) - 12 out of 16 first-place rankings on depth completion benchmarks - RMSE cut in half: 0.132 → 0.062 on the hardest indoor scenes LingBot-Vision trained on boundaries, because object edges carry the geometry robots need. No human boundary labels are used, which makes this approach easier to scale. The open-sourced LingBot-Vision is the general vision backbone, and LingBot-Depth 2.0 is the depth model built on it.

The robot’s “eyes” just received a big upgrade. LingBot-Depth 2.0, a depth-completion model with half the depth error just dropped. 12/16 benchmarks topped. Glass, mirrors, and transparent objects are so easy for us humans, but so hard for robots, because they do not behave like ordinary surfaces in a camera pipeline. A robot that misunderstands a balcony window or a table edge, will have a completely false planning inside a false world. Huge implecation. LingBot-Depth 2.0 takes an RGB image plus a broken depth map from a sensor and then outputs a cleaner depth map and a usable 3D point cloud. Numbers on LingBot-Depth 2.0 - Excels on glass, mirrors & transparent objects — where traditional depth cameras fail - Training data: 3M → 150M (50x scale-up) - 12 out of 16 first-place rankings on depth completion benchmarks - RMSE cut in half: 0.132 → 0.062 on the hardest indoor scenes LingBot-Vision trained on boundaries, because object edges carry the geometry robots need. No human boundary labels are used, which makes this approach easier to scale. The open-sourced LingBot-Vision is the general vision backbone, and LingBot-Depth 2.0 is the depth model built on it.

Rohan Paul

25,312 次观看 • 11 天前

🗣️ "It's the biggest win of my career by far." Arno Merk insists he can pose a threat to 'anybody' after dumping out two-time World Champion Peter Wright in straight sets! #WCDarts

🗣️ "It's the biggest win of my career by far." Arno Merk insists he can pose a threat to 'anybody' after dumping out two-time World Champion Peter Wright in straight sets! #WCDarts

PDC Darts

27,420 次观看 • 6 个月前

🚨Jeffrey Sachs: "It's time for India, China, Russia, and others to tell the United States publicly as well as privately, back off. 🚨You don't run the world and you can't wreck the world. 🚨And especially to tell pipsqueak Israel, stop wrecking the whole world out of your visions of your power."

🚨Jeffrey Sachs: "It's time for India, China, Russia, and others to tell the United States publicly as well as privately, back off. 🚨You don't run the world and you can't wreck the world. 🚨And especially to tell pipsqueak Israel, stop wrecking the whole world out of your visions of your power."

Ignorance, the root and stem of all evil

36,273 次观看 • 3 个月前

World models are moving beyond offline generation towards interactive, real-time experiences. Introducing ⚡FlashDreams⚡: an open-source high-performance inference and serving library built for autoregressive world models: 🔥 Up to 3.10× faster LingBot-World inference 🔥 Up to 2.12× faster Self-Forcing inference 🔥 Up to 1.40× faster Wan2.1 inference 🔥 8 integrated models 🔥 Multi-GPU, streaming, low-latency serving 🔥 Agentic skills that teach you how to use it FlashDreams is designed for a new generation of AI systems that continuously evolve over time while responding to user interactions. It powers applications across robotics, autonomous vehicle simulation, gaming, and virtual worlds. Github: Docs: Research page: Join the #flashdreams Discord channel at FlashDreams is also the runtime backbone behind NVIDIA OmniDreams ( 1/n #AI #WorldModels #FastInference #PhysicalAI #OpenSource #NVIDIA

World models are moving beyond offline generation towards interactive, real-time experiences. Introducing ⚡FlashDreams⚡: an open-source high-performance inference and serving library built for autoregressive world models: 🔥 Up to 3.10× faster LingBot-World inference 🔥 Up to 2.12× faster Self-Forcing inference 🔥 Up to 1.40× faster Wan2.1 inference 🔥 8 integrated models 🔥 Multi-GPU, streaming, low-latency serving 🔥 Agentic skills that teach you how to use it FlashDreams is designed for a new generation of AI systems that continuously evolve over time while responding to user interactions. It powers applications across robotics, autonomous vehicle simulation, gaming, and virtual worlds. Github: Docs: Research page: Join the #flashdreams Discord channel at FlashDreams is also the runtime backbone behind NVIDIA OmniDreams ( 1/n #AI #WorldModels #FastInference #PhysicalAI #OpenSource #NVIDIA

Ruilong Li

90,235 次观看 • 1 个月前

.Balaji says AI verification is much harder in the digital world than it is for the physical world: "AI for the physical world is very verifiable because the thing is, the digital world is fundamentally decentralized in a way the physical world isn't." "There's only one physical world. You can say: did the AI move this box from this pallet to that pallet? That is something where you can probably get it to 100% over time." "By contrast, in the digital world, there are all these people who live in their own constructed environments." "How do you know when you're done with your to-do list? That's harder. Those things are fuzzier."

.Balaji says AI verification is much harder in the digital world than it is for the physical world: "AI for the physical world is very verifiable because the thing is, the digital world is fundamentally decentralized in a way the physical world isn't." "There's only one physical world. You can say: did the AI move this box from this pallet to that pallet? That is something where you can probably get it to 100% over time." "By contrast, in the digital world, there are all these people who live in their own constructed environments." "How do you know when you're done with your to-do list? That's harder. Those things are fuzzier."

a16z

132,718 次观看 • 3 个月前

Introducing Modality Forcing, a recipe for post-training T2I models for SOTA RGB-Depth generation! Text-to-image (T2I) models learn rich representations of the spatial world. How do we build on this prior for high-quality depth generation? 🧵 [1/6]

Introducing Modality Forcing, a recipe for post-training T2I models for SOTA RGB-Depth generation! Text-to-image (T2I) models learn rich representations of the spatial world. How do we build on this prior for high-quality depth generation? 🧵 [1/6]

Bardienus Duisterhof

64,543 次观看 • 1 个月前

🔥 World of Dypians delivers some of the best graphics in Web3 gaming. High-quality environments, detailed characters, and a living open world that sets a new visual standard for blockchain games. The world feels alive, the atmosphere is immersive, and every area is crafted with AAA-level attention to detail, making the experience stand out compared to anything else in the space.

🔥 World of Dypians delivers some of the best graphics in Web3 gaming. High-quality environments, detailed characters, and a living open world that sets a new visual standard for blockchain games. The world feels alive, the atmosphere is immersive, and every area is crafted with AAA-level attention to detail, making the experience stand out compared to anything else in the space.

World of Dypians

200,274 次观看 • 7 个月前