Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Video understanding isn't just recognizing —it demands reasoning across thousands of frames. Meet Long-RL🚀 Highlights: 🧠 Dataset: LongVideo-Reason — 52K QAs with reasoning. ⚡ System: MR-SP - 2.1× faster RL for long videos. 📈 Scalability: Hour-long videos (3,600 frames) RL on a single node (8×A100s). 🖼️📝🎵 RL training for... show more

Yukang Chen

1,488 subscribers

31,652 views • 11 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

🚀 We open-sourced LongLive — interactive, real-time long-video generation. 👥Generates video in real time as users enter text prompts. ⚡️20.7 FPS on a single H100,⏱️up to 240s per clip. 🎬Fine-tunes SOTA short-video models (e.g., Wan) into long-video generators. 🌍One step closer to World Models. All code for training & inference, model weights, demo page, and videos released! Paper: Code: Model: Demo Page: Introduction Video:

🚀 We open-sourced LongLive — interactive, real-time long-video generation. 👥Generates video in real time as users enter text prompts. ⚡️20.7 FPS on a single H100,⏱️up to 240s per clip. 🎬Fine-tunes SOTA short-video models (e.g., Wan) into long-video generators. 🌍One step closer to World Models. All code for training & inference, model weights, demo page, and videos released! Paper: Code: Model: Demo Page: Introduction Video:

Yukang Chen

11,752 views • 8 months ago

verifiers v0.1.7 is released 🚀 this one's all about making RL training and experimentation waaaay easier: - single-command installation for prime-rl - single-command training w/ unified configs - overhauled vf.RLTrainer for hacking on new algorithms quick demo + links below :)

verifiers v0.1.7 is released 🚀 this one's all about making RL training and experimentation waaaay easier: - single-command installation for prime-rl - single-command training w/ unified configs - overhauled vf.RLTrainer for hacking on new algorithms quick demo + links below :)

will brown

28,579 views • 7 months ago

ByteDance just unveiled HuMo, a unified framework for human-centric video generation. With HuMo you can: • Generate videos from text + image • Create audio-synced videos from text + audio • Combine text, image, and audio for maximum control • Preserve consistent subjects across frames • Achieve natural audio-visual synchronization This isn’t just another release , it’s a big step toward fully controllable, coherent, and cinematic AI video.

ByteDance just unveiled HuMo, a unified framework for human-centric video generation. With HuMo you can: • Generate videos from text + image • Create audio-synced videos from text + audio • Combine text, image, and audio for maximum control • Preserve consistent subjects across frames • Achieve natural audio-visual synchronization This isn’t just another release , it’s a big step toward fully controllable, coherent, and cinematic AI video.

DStudioproject

29,408 views • 9 months ago

We open-sourced QeRL — Quantization-enhanced Reinforcement Learning ! 🧠 4-bit quantized RL training 💪 Train a 32B LLM on a single H100 GPU ⚙️ 1.7× faster overall training 🎯 Accuracy on par with bfloat16-level accuracy 🔥 Supports NVFP4 quantization format Moreover, we show that quantization helps exploration in RL training. Paper: Code: #NVIDIA #AIResearch #ReinforcementLearning #Quantization #LLM #EfficientAI

We open-sourced QeRL — Quantization-enhanced Reinforcement Learning ! 🧠 4-bit quantized RL training 💪 Train a 32B LLM on a single H100 GPU ⚙️ 1.7× faster overall training 🎯 Accuracy on par with bfloat16-level accuracy 🔥 Supports NVFP4 quantization format Moreover, we show that quantization helps exploration in RL training. Paper: Code: #NVIDIA #AIResearch #ReinforcementLearning #Quantization #LLM #EfficientAI

Yukang Chen

69,720 views • 8 months ago

Today we’re launching INTELLECT-2: The first decentralized 32B-parameter RL training run open to join for anyone with compute — fully permissionless. Scaling towards frontier reasoning across coding, math and science.

Today we’re launching INTELLECT-2: The first decentralized 32B-parameter RL training run open to join for anyone with compute — fully permissionless. Scaling towards frontier reasoning across coding, math and science.

Prime Intellect

353,394 views • 1 year ago

We developed an RL method for fine-tuning our models for precise tasks in just a few hours or even minutes. Instead of training the whole model, we add an “RL token” output to π-0.6, our latest model, which is used by a tiny actor and critic to learn quickly with RL.

We developed an RL method for fine-tuning our models for precise tasks in just a few hours or even minutes. Instead of training the whole model, we add an “RL token” output to π-0.6, our latest model, which is used by a tiny actor and critic to learn quickly with RL.

Physical Intelligence

429,829 views • 2 months ago

Introducing INTELLECT-3: Scaling RL to a 100B+ MoE model on our end-to-end stack Achieving state-of-the-art performance for its size across math, code and reasoning Built using the same tools we put in your hands, from environments & evals, RL frameworks, sandboxes & more

Introducing INTELLECT-3: Scaling RL to a 100B+ MoE model on our end-to-end stack Achieving state-of-the-art performance for its size across math, code and reasoning Built using the same tools we put in your hands, from environments & evals, RL frameworks, sandboxes & more

Prime Intellect

1,133,644 views • 6 months ago

Hiring RL Engineer! Started off as a curious project at Lossfunk to push the boundaries of LLMs in social reasoning - we are now building RL environments, data, and benchmarks to simulate more real-world scenarios. If you want to train SoTA RL models over multi-GPUs (H200s/B200s) to unlock next AI frontier, this is for you.

Hiring RL Engineer! Started off as a curious project at Lossfunk to push the boundaries of LLMs in social reasoning - we are now building RL environments, data, and benchmarks to simulate more real-world scenarios. If you want to train SoTA RL models over multi-GPUs (H200s/B200s) to unlock next AI frontier, this is for you.

Satpal Singh Rathore

45,915 views • 10 months ago

Introducing Reinforcement-Learned Teachers (RLTs): Transforming how we teach LLMs to reason with reinforcement learning (RL). Blog: Paper: Traditional RL focuses on “learning to solve” challenging problems with expensive LLMs and constitutes a key step in making student AI systems ultimately acquire reasoning capabilities via distillation and cold-starting. Enter our RLTs—a new class of models prompted with not only a problem’s question but also its solution, and directly trained to generate clear, step-by-step “explanations” to teach their students. Remarkably, an RLT with only 7B parameters produces superior results when distilling and cold-starting students in competitive and graduate-level reasoning tasks than orders-of-magnitude larger LLMs. RLTs are as effective even when distilling 32B students, much larger than the teacher itself—unlocking a new standard for efficiency in developing reasoning language models with RL. Code:

Introducing Reinforcement-Learned Teachers (RLTs): Transforming how we teach LLMs to reason with reinforcement learning (RL). Blog: Paper: Traditional RL focuses on “learning to solve” challenging problems with expensive LLMs and constitutes a key step in making student AI systems ultimately acquire reasoning capabilities via distillation and cold-starting. Enter our RLTs—a new class of models prompted with not only a problem’s question but also its solution, and directly trained to generate clear, step-by-step “explanations” to teach their students. Remarkably, an RLT with only 7B parameters produces superior results when distilling and cold-starting students in competitive and graduate-level reasoning tasks than orders-of-magnitude larger LLMs. RLTs are as effective even when distilling 32B students, much larger than the teacher itself—unlocking a new standard for efficiency in developing reasoning language models with RL. Code:

Sakana AI

179,219 views • 11 months ago

The 13th and final RL GRIME Halloween mix (hurts to type that) is out 👑 RL got Jack Black for the intro and it starts with a new RL GRIME x WINK collab SABLE VALLEY

The 13th and final RL GRIME Halloween mix (hurts to type that) is out 👑 RL got Jack Black for the intro and it starts with a new RL GRIME x WINK collab SABLE VALLEY

Dancing Astronaut

19,815 views • 1 year ago

🕹️We are excited to introduce "ChronoEdit: Towards Temporal Reasoning for Image Editing and World Simulation" ChronoEdit reframes image editing as a video generation task to encourage temporal consistency. It leverages a temporal reasoning stage that denoises with “video reasoning tokens” to "reason" on physically plausible edits. See the attached video for results. Project Page: Arxiv: Code and model are coming.

🕹️We are excited to introduce "ChronoEdit: Towards Temporal Reasoning for Image Editing and World Simulation" ChronoEdit reframes image editing as a video generation task to encourage temporal consistency. It leverages a temporal reasoning stage that denoises with “video reasoning tokens” to "reason" on physically plausible edits. See the attached video for results. Project Page: Arxiv: Code and model are coming.

Huan Ling

36,841 views • 8 months ago

🚨 LTX 2.3 Quality is now live on fal! 🎬 New Higher Quality LTX 2.3 video generation ⚡ Text, image, audio & reference-video workflows 🎛️ LoRA variants for every endpoint 🌈 HDR with EXR frames + MP4 preview

🚨 LTX 2.3 Quality is now live on fal! 🎬 New Higher Quality LTX 2.3 video generation ⚡ Text, image, audio & reference-video workflows 🎛️ LoRA variants for every endpoint 🌈 HDR with EXR frames + MP4 preview

fal

20,364 views • 14 days ago

We asked Sholto Douglas from Anthropic about the costs of RL (Reinforcement Learning) runs. "In Dario Amodei's essay, he said that RL runs cost only $1M back in December." "RL is a more naively parallelizable and scalable than pre-training." "With pre-training, you need everything in one big data center ideally. For RL, in theory, you could scale all over the world."

We asked Sholto Douglas from Anthropic about the costs of RL (Reinforcement Learning) runs. "In Dario Amodei's essay, he said that RL runs cost only $1M back in December." "RL is a more naively parallelizable and scalable than pre-training." "With pre-training, you need everything in one big data center ideally. For RL, in theory, you could scale all over the world."

TBPN

76,634 views • 1 year ago

🚀 Introducing T* and LV-Haystack — our latest leap forward in VLMs for long video understanding! 🧩 Lightweight plugin: T* boosting LLaVA-OV-72B (56→62%) and GPT-4o (50→53%)! ⚡ Fast inference: 34.9s → 10.4s latency, 691 → 170 TFLOPs v.s. SOTA. 📚 Large-scale dataset: 400 hours of videos + 15,000 samples. → TLDR: Long video understanding needs fine-grained supervision, so we start with exploring "key frames" as the first step 🪜. 1/

🚀 Introducing T* and LV-Haystack — our latest leap forward in VLMs for long video understanding! 🧩 Lightweight plugin: T* boosting LLaVA-OV-72B (56→62%) and GPT-4o (50→53%)! ⚡ Fast inference: 34.9s → 10.4s latency, 691 → 170 TFLOPs v.s. SOTA. 📚 Large-scale dataset: 400 hours of videos + 15,000 samples. → TLDR: Long video understanding needs fine-grained supervision, so we start with exploring "key frames" as the first step 🪜. 1/

Zihan "Zenus" Wang ✈️ ICLR

49,520 views • 1 year ago

Launching SYNTHETIC-2: our next-gen open reasoning dataset and planetary-scale synthetic data generation run. Powered by our P2P inference stack and DeepSeek-R1-0528, it verifies traces for the hardest RL tasks. Contribute towards AGI via open, permissionless compute.

Launching SYNTHETIC-2: our next-gen open reasoning dataset and planetary-scale synthetic data generation run. Powered by our P2P inference stack and DeepSeek-R1-0528, it verifies traces for the hardest RL tasks. Contribute towards AGI via open, permissionless compute.

Prime Intellect

359,732 views • 11 months ago

🤖 New paper: MobileVLA-R1 A unified VLA system that brings real reasoning + continuous control to quadruped robots. CoT dataset, 2-stage training, real-world deployment. 📄paper & code & demo:

🤖 New paper: MobileVLA-R1 A unified VLA system that brings real reasoning + continuous control to quadruped robots. CoT dataset, 2-stage training, real-world deployment. 📄paper & code & demo:

Hao Tang (hiring postdocs)

27,232 views • 6 months ago

We developed a simple, sample-efficient online RL technique for post-training image generation models. We see it as a possible steerable alternative to CFG, driven by any scalar reward, including human preference.

We developed a simple, sample-efficient online RL technique for post-training image generation models. We see it as a possible steerable alternative to CFG, driven by any scalar reward, including human preference.

David McAllister

63,671 views • 1 month ago

Meet LongCat-Video-Avatar: a robust audio-driven avatar model that pushes the boundaries of long-form video generation. Compared with the previous InfiniteTalk, LongCat-Video-Avatar delivers far better long-sequence stability and realism. New highlights： ⚙ Built on the LongCat-Video architecture, now supporting Audio-Text-to-Video (AT2V), Audio-Text-Image-to-Video (ATI2V), and Video Continuation modes. 🎭 Open-source SOTA Realism: Ranked 1st in overall anthropomorphism scores for both single and multi-subject scenarios in EvalTalker evaluations (492 participants, 3 independent raters per video). ♾ High-quality long videos: Cross-Chunk Latent Stitching prevents pixel degradation and error accumulation over time, ensuring seamless stitching quality. 🔒 Long-term consistency: Reference Skip Attention maintains ID consistency while eliminating rigid copy-paste artifacts. 🪄 Supports multi-person and infinite-length video generation. 🔗Open-sourced Code: Hugging Face: Project: Paper:

Meituan LongCat

28,020 views • 6 months ago

🚀 Excited to release LongLive 2.0! 🎬 An end-to-end infrastructure for long video generation, with FP4 and parallelism at the core of both training and inference. ⚡45.7 FPS generation speed on 5B model⚡ ✨ LongLive 2.0 supports real-video training, few-step distillation, multi-shot training/inference, sequence-parallel acceleration, NVFP4 KV cache, and async VAE decoding deployment. 🧩 To our knowledge, this is the first open-source 4-bit long video generation infra that covers both training and inference. 🙌 Welcome to check it out, try it, and share feedback! 🔗 Code: 📰 Paper: 🎥 Demo: #LongVideoGeneration #VideoGeneration #Realtime #AIInfra #EfficientAI #FP4 #Parallel #NVIDIA

🚀 Excited to release LongLive 2.0! 🎬 An end-to-end infrastructure for long video generation, with FP4 and parallelism at the core of both training and inference. ⚡45.7 FPS generation speed on 5B model⚡ ✨ LongLive 2.0 supports real-video training, few-step distillation, multi-shot training/inference, sequence-parallel acceleration, NVFP4 KV cache, and async VAE decoding deployment. 🧩 To our knowledge, this is the first open-source 4-bit long video generation infra that covers both training and inference. 🙌 Welcome to check it out, try it, and share feedback! 🔗 Code: 📰 Paper: 🎥 Demo: #LongVideoGeneration #VideoGeneration #Realtime #AIInfra #EfficientAI #FP4 #Parallel #NVIDIA

Yukang Chen

57,091 views • 27 days ago

.Polymath is training world generation models to automate the creation of RL environments. Traditionally, RL environment generation has been bottlenecked by human data. Superintelligence will never be achieved by human data alone. Polymath is building the core technology to enable automated environment generation using far less human effort than traditionally required, and eventually none. This allows for more complex and realistic worlds, and higher quality, scale, and diversity of tasks. This will be essential to unlock RL scaling. The end goal is to create large-scale, long-horizon environments from a text description alone. This will enable the creation of worlds of arbitrary complexity and scale, which is foundational for training & evaluating autonomous, superintelligent AI agents. Congrats on the launch, Dylan Ma and Naren Yenuganti!

.Polymath is training world generation models to automate the creation of RL environments. Traditionally, RL environment generation has been bottlenecked by human data. Superintelligence will never be achieved by human data alone. Polymath is building the core technology to enable automated environment generation using far less human effort than traditionally required, and eventually none. This allows for more complex and realistic worlds, and higher quality, scale, and diversity of tasks. This will be essential to unlock RL scaling. The end goal is to create large-scale, long-horizon environments from a text description alone. This will enable the creation of worlds of arbitrary complexity and scale, which is foundational for training & evaluating autonomous, superintelligent AI agents. Congrats on the launch, Dylan Ma and Naren Yenuganti!

Y Combinator

44,192 views • 3 months ago