Loading video...

Video Failed to Load

Go Home

[1/N]🚀New decoding paradigm drop!🚀 Introducing Lookahead Reasoning(LR): step-level speculation that stacks with Speculative Decoding(SD). It has been accepted to #NeurIPS2025 🎉 📖 Blog: 💻 Code: 📄 Paper:

haoailab's profile picture

Hao AI Lab

6,640 subscribers

43,199 views • 1 year ago •via X (Twitter)

EducationScience & Technology#NeurIPS2025
LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

We’re open-sourcing lithos-metal 🚀 Megakernels + DSpark speculative decoding run Qwen3.8-27B at 200+ tokens/s/user peak on one Apple M5 Max. Ultra-fast inference on your laptop. Try it with any coding agent in one command. Code: Tech blog:
0:37

We’re open-sourcing lithos-metal 🚀 Megakernels + DSpark speculative decoding run Qwen3.8-27B at 200+ tokens/s/user peak on one Apple M5 Max. Ultra-fast inference on your laptop. Try it with any coding agent in one command. Code: Tech blog:

Zhihao Jia

423,198 views • 27 days ago

Introducing MiniCPM 4.1-8B: First Open-Source Reasoning LLM with Trainable Sparse Attention ✅ Strong Reasoning Capability: Surpasses similar-sized models on 15 tasks! ✅ Fast Generation: 3x decoding speedup for reasoning ✅ Efficient Architecture: Trainable sparse attention, frequency-ranked speculative decoding Download Models: Huggingface: Github: Technical Report: #AI #MiniCPM #LLM #OpenBMB #ArtificialIntelligence #MachineLearning
0:54

Introducing MiniCPM 4.1-8B: First Open-Source Reasoning LLM with Trainable Sparse Attention ✅ Strong Reasoning Capability: Surpasses similar-sized models on 15 tasks! ✅ Fast Generation: 3x decoding speedup for reasoning ✅ Efficient Architecture: Trainable sparse attention, frequency-ranked speculative decoding Download Models: Huggingface: Github: Technical Report: #AI #MiniCPM #LLM #OpenBMB #ArtificialIntelligence #MachineLearning

OpenBMB

19,317 views • 1 year ago

🚀 Introducing #SparseVideoGen: 2x speedup in video generation with HunyuanVideo with high pixel-level fidelity (PSNR = 29)! No training is required, no perceptible difference to the human eye! Blog: Paper: Code: #ArtificialIntelligence #OpenSource [1/7]
0:46

🚀 Introducing #SparseVideoGen: 2x speedup in video generation with HunyuanVideo with high pixel-level fidelity (PSNR = 29)! No training is required, no perceptible difference to the human eye! Blog: Paper: Code: #ArtificialIntelligence #OpenSource [1/7]

Haocheng Xi

45,770 views • 1 year ago

🎉 MiniMax M3.1 Flash from MiniMax (official) has Day-0 support in vLLM! Built for multi-step automation, real-codebase engineering and visual reasoning—with native NVFP4 KV, NVFP4 MoE weights and DSpark speculative decoding in the serving stack. What the model can do, and the vLLM engineering that makes it practical to serve. 👇 Model: 1/5
0:40

🎉 MiniMax M3.1 Flash from MiniMax (official) has Day-0 support in vLLM! Built for multi-step automation, real-codebase engineering and visual reasoning—with native NVFP4 KV, NVFP4 MoE weights and DSpark speculative decoding in the serving stack. What the model can do, and the vLLM engineering that makes it practical to serve. 👇 Model: 1/5

vLLM

11,710 views • 13 days ago

🚀 Big news from Recursion! Our 2nd Neuroscience Phenomap — a whole-genome map of the brain’s immune cells 🧠 — has been accepted by Roche & Genentech, triggering a $30M milestone! A massive step toward decoding biology to radically improve lives. 💥 #TechBio #AI #Neuroscience
0:35

🚀 Big news from Recursion! Our 2nd Neuroscience Phenomap — a whole-genome map of the brain’s immune cells 🧠 — has been accepted by Roche & Genentech, triggering a $30M milestone! A massive step toward decoding biology to radically improve lives. 💥 #TechBio #AI #Neuroscience

Chris Gibson

25,535 views • 1 year ago

🎬 VChain is accepted to #ACL2026 Findings! 🚀 #VChain is an inference-time framework that injects reasoning signals (*Chain-of-Visual-Thought*) from MLLMs to guide video generation for better consistency. 🔗 Website: 📄 Paper:
2:00

🎬 VChain is accepted to #ACL2026 Findings! 🚀 #VChain is an inference-time framework that injects reasoning signals (*Chain-of-Visual-Thought*) from MLLMs to guide video generation for better consistency. 🔗 Website: 📄 Paper:

Ziqi Huang

13,774 views • 6 months ago

🚀 Our DSpark for Qwen3.8-27B beats native MTP with the same 8 speculative tokens on 4×H200. Up to 52% faster single-stream decoding and 23% higher peak throughput. On 8-needle MRCR, it averages 4.30 accepted tokens beyond 1M context. Model: Demo below, see for yourself!
0:54

🚀 Our DSpark for Qwen3.8-27B beats native MTP with the same 8 speculative tokens on 4×H200. Up to 52% faster single-stream decoding and 23% higher peak throughput. On 8-needle MRCR, it averages 4.30 accepted tokens beyond 1M context. Model: Demo below, see for yourself!

Red Hat AI

14,041 views • 1 month ago

Steps to watch this video?👇🤩 Step 1 - Video is only for fun.. No Decoding, Superfans! 🥳 Step 2 - Video is only for fun.. No Decoding, Superfans! 🥳 Step 3 - Video is only for fun.. No Decoding, Superfans!🥳 #WhistlePodu with Tamil Nadu Weatherman
2:06

Steps to watch this video?👇🤩 Step 1 - Video is only for fun.. No Decoding, Superfans! 🥳 Step 2 - Video is only for fun.. No Decoding, Superfans! 🥳 Step 3 - Video is only for fun.. No Decoding, Superfans!🥳 #WhistlePodu with Tamil Nadu Weatherman

Chennai Super Kings

1,070,979 views • 21 hours ago

🚀 Introducing Zed v0.159! Today we're excited to launch Zed SSH Remoting in Beta. 💻🌐🎉 See the blog post for more details.
1:02

🚀 Introducing Zed v0.159! Today we're excited to launch Zed SSH Remoting in Beta. 💻🌐🎉 See the blog post for more details.

Zed

57,094 views • 2 years ago

Introducing: the new llama-cli 🦙🦙 > Clean looking interface > Multimodal support > Conversation control via commands > Speculative decoding support > Jinja fully supported
0:26

Introducing: the new llama-cli 🦙🦙 > Clean looking interface > Multimodal support > Conversation control via commands > Speculative decoding support > Jinja fully supported

Xuan-Son Nguyen

21,343 views • 10 months ago

Decoding Morse Code.
0:40

Decoding Morse Code.

Interesting STEM

32,753 views • 1 year ago

🚀Create complex, modular 3D assets and scenes.🚀 Introducing #MoCA (Mixture-of-Components Attention), a new sparse attention method for compositional 3D generation. It scales to *many * components. paper: code:
0:31

🚀Create complex, modular 3D assets and scenes.🚀 Introducing #MoCA (Mixture-of-Components Attention), a new sparse attention method for compositional 3D generation. It scales to *many * components. paper: code:

Tengfei Wang

27,212 views • 11 months ago

Fast Byte Latent Transformer is accepted to ICML 2026! ⚡🥪 Byte-level LMs promise to free us from subword tokenizers, but decoding one byte at a time is super slow. We make BLT generation more efficient with BLT-D: text diffusion for parallel byte decoding. 1/
1:46

Fast Byte Latent Transformer is accepted to ICML 2026! ⚡🥪 Byte-level LMs promise to free us from subword tokenizers, but decoding one byte at a time is super slow. We make BLT generation more efficient with BLT-D: text diffusion for parallel byte decoding. 1/

Julie Kallini✨COLM26

107,444 views • 5 months ago

While exploring diffusion language models (DLMs), we kept asking ourselves: given speculative decoding, a lossless acceleration technique, why do we still need DLMs? 🤔 Our latest work, ORL, tries to answer this question: AR optimizes intelligence 👆, speculative decoding optimizes speed 👉, and DLMs optimize the Pareto frontier of speed and intelligence 🚀. Through co-design across the algorithms, the training and inference engine, and the kernels, ORL pushes out a new Pareto frontier of intelligence and speed. Similar performance, much faster than speculative decoding. You say lossy spec? ORL-DLM is better too! 😎 🌐 Project + live demos:
1:28

While exploring diffusion language models (DLMs), we kept asking ourselves: given speculative decoding, a lossless acceleration technique, why do we still need DLMs? 🤔 Our latest work, ORL, tries to answer this question: AR optimizes intelligence 👆, speculative decoding optimizes speed 👉, and DLMs optimize the Pareto frontier of speed and intelligence 🚀. Through co-design across the algorithms, the training and inference engine, and the kernels, ORL pushes out a new Pareto frontier of intelligence and speed. Similar performance, much faster than speculative decoding. You say lossy spec? ORL-DLM is better too! 😎 🌐 Project + live demos:

Chenfeng_X

21,273 views • 20 days ago

I’m excited to announce that 💫StarVector has been accepted at CVPR 2025! Over a year in the making, StarVector opens a new paradigm for Scalable Vector Graphics (SVG) generation by harnessing multimodal LLMs to generate SVG code that aesthetically mirrors input images and text. With this milestone, we’re also releasing StarVector on Hugging Face! 🥳🚀
0:55

I’m excited to announce that 💫StarVector has been accepted at CVPR 2025! Over a year in the making, StarVector opens a new paradigm for Scalable Vector Graphics (SVG) generation by harnessing multimodal LLMs to generate SVG code that aesthetically mirrors input images and text. With this milestone, we’re also releasing StarVector on Hugging Face! 🥳🚀

Joan Rodriguez

16,410 views • 1 year 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:
3:48

🤖 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)

27,306 views • 11 months ago

Code Micky - Momo Yaa |(Decoding)
7:11

Code Micky - Momo Yaa |(Decoding)

code micky

188,709 views • 2 years ago

This Decoding.. code micky #ForNothing
5:12

This Decoding.. code micky #ForNothing

Kofi Kinaata

16,317 views • 6 months ago