JD Web Programmer
@jdwebprogrammer
Followers
23
Following
823
Media
912
Statuses
1K
Solving AGI
Joined February 2021
How big would LLM's be if each neuron was a grain of sand? 🔶 100B ≈ a cube of sand ~4.6 m on a side (since 4.6³ ≈ 100). 🔶 500B ≈ cube ~7.9 m per side. 🔶 1T ≈ cube ~10 m per side. 🔶 2T ≈ cube ~12.6 m per side.
0
0
0
📜 Copy and use the generated code script 📦 Download the model file directly ⚡ Animate your custom model with live inference! 🔗 https://t.co/Vnpdxsv7W3
1
0
0
🚀 Build neural networks in a 3D interface! 🧵👇 📚 Comes with model presets for learning 💎 Design your own architecture visually
2
0
1
⚡ Faster than Fast. Designed for Agentic AI. Introducing Xiaomi MiMo-V2-Flash — our new open-source MoE model: 309B total params, 15B active. Blazing speed meets frontier performance. 🔥 Highlights: 🏗️ Hybrid Attention: 5:1 interleaved 128-window SWA + Global | 256K context 📈
90
303
2K
“See Once, Then Act” shows robots can learn manipulation tasks from a single human demo video. The ViVLA model watches an expert video and maps it to robot actions, enabling one-shot task learning without retraining.
2
27
110
💫 "Demultiplexing through a multimode fiber using chip-scale diffractive neural network" *Light speed neural networking* The authors present the first fully optical, chip-scale mode demultiplexer for multimode fibers (MMFs) based on a 3D-nanoprinted diffractive neural network
0
0
1
💫 "Structured Light at the Extreme: Harnessing Spatiotemporal Control for High-Field Laser-Matter Interactions" 🚨 Structured light and optics has just been advanced! 🚀 The next generation in high-intensity laser physics will not come from higher power alone, but from precise,
0
0
1
📑 "4DLangVGGT: 4D Language-Visual Geometry Grounded Transformer" The authors propose 4DLangVGGT, the first feed-forward, Transformer-based framework that jointly models 4D geometry and open-vocabulary language semantics in a single unified network, eliminating the need for
0
0
1
💎 The Geometry of Intelligence 🧠 🗒️ This new very interesting paper shows that, for any real physical signal, with only a small set of traces you can discover the complete compact perceptual manifold. You find it by tracking the Hausdorff radius (the running maximum
0
0
1
New paper 📜: Tiny Recursion Model (TRM) is a recursive reasoning approach with a tiny 7M parameters neural network that obtains 45% on ARC-AGI-1 and 8% on ARC-AGI-2, beating most LLMs. Blog: https://t.co/w5ZDsHDDPE Code: https://t.co/7UgKuD9Yll Paper:
arxiv.org
Hierarchical Reasoning Model (HRM) is a novel approach using two small neural networks recursing at different frequencies. This biologically inspired method beats Large Language models (LLMs) on...
146
671
4K
Check out our latest work: a dexterous hand that can actually turn screws. And yes, our code for RL sim-to-real, lightweight BC, and skill-based teleop is all open-sourced! Project page: https://t.co/2drK5KPSbM RL sim2real: https://t.co/5AfVTuWzGs MinBC:
github.com
MinBC - Minimal Behavior Cloning. Contribute to x-robotics-lab/minbc development by creating an account on GitHub.
How do we get a robot to use a screwdriver 🪛 and fasten a nut 🔩 ? Introducing DexScrew, our sim-to-real framework that enables a dexterous hand 🖐️ to autonomously perform complex and contact-rich tasks even when we cannot accurately simulate them. We open-source the
5
22
63
A 2-hour uncut video of X-VLA folding cloth at a 100% success rate. we provide the checkpoints, and you can easily fine-tune them on any dataset (any action dimension): https://t.co/kU0Stfkoq8 enjoy!
🚀 Introducing X-VLA ; LeRobot’s new soft-prompted Vision-Language-Action model. X-VLA is built to scale across many embodiments: different robots, cameras, action spaces, and environments, all handled by one unified transformer backbone. - Generalist across robots (Franka,
7
35
269
We just solved quantum tunneling! What's next this week?
Call it whatever you want, but when a person phases through a wall, that's teleportation to me. Macroscopic quantum tunneling just won the Nobel Prize in Physics.
0
0
0
All ingredients for 'RNA World' found on asteroid Bennu. The samples contain presolar grains (silicon carbide, graphite) that formed around other stars before the Sun existed (up to 7 billion years old).
BREAKING: Sugars essential for life have been found in pristine asteroid Bennu samples collected by NASA’s OSIRIS-REx spacecraft. Combined with previous detections of amino acids and nucleobases, we see that life’s ingredients were widespread throughout the solar system:
0
0
0
Classify intents for better mobile agents
0
0
0
📑 "Dynamic LSTM-based Memory Encoder For Long-term LLM Interactions" This interesting paper uses an LSTM as memory for LLMs but used only for user preferences. Although could this be a key step toward an #AGI architecture? 🤖 Grok summary of the paper: The authors propose
0
0
1
📑 "Dynamic LSTM-based Memory Encoder For Long-term LLM Interactions" This interesting paper uses an LSTM as memory for LLMs but used only for user preferences. Although could this be a key step toward an #AGI architecture? 🤖 Grok summary of the paper: The authors propose
0
0
0