#mujoco
LightNews LightNews
kdaic.bsky.social
mujoco 4.0
x.com/mujoco/statu...
以下のリンクで遊べる。たのしい
mujoco.org
MuJoCo — Advanced Physics Simulation
MuJoCo — Advanced Physics Simulation. Interactive MuJoCo 4.0 physics showcase running natively in your browser via WebAssembly.
mujoco.org
November 9, 2026 at 2:55 PM
deandc43.bsky.social
MuJoCo 4
A physics engine for robots with its own studio Discussion | Link
dlvr.it
November 10, 2026 at 8:01 AM
k1imber.bsky.social
💫 Физический симулятор MuJoCo для исследований и разработки

MuJoCo — физический движок с открытым исходным кодом для моделирования взаимодействий сложных структур с окружающей средой. Поддерживает C API и Python-бины.

• C API для разработчиков
• Интерактивная визуализация в браузере и десктопе
• …
November 10, 2026 at 7:00 AM
github-trending.bsky.social
📦 google-deepmind / mujoco
⭐ 16,148 (+26)
🗒 C++

Multi-Joint dynamics with Contact. A general purpose physics simulator.
GitHub - google-deepmind/mujoco: Multi-Joint dynamics with Contact. A general purpose physics simulator.
Multi-Joint dynamics with Contact. A general purpose physics simulator. - google-deepmind/mujoco
github.com
November 8, 2026 at 6:17 PM
dailygithubtrends.bsky.social
今日のGitHubトレンド

google-deepmind/mujoco
MuJoCoは、ロボティクスや機械学習などの研究開発に向けた汎用物理エンジンです。
関節構造と環境の相互作用を高速かつ正確にシミュレートすることを目的としています。
C APIによる高性能な動作に加え、可視化機能やPythonバインディング、Unityプラグインなどを提供しています。
GitHub - google-deepmind/mujoco: Multi-Joint dynamics with Contact. A general purpose physics simulator. · GitHub
Multi-Joint dynamics with Contact. A general purpose physics simulator. - google-deepmind/mujoco
github.com
November 9, 2026 at 9:21 PM
shei-designs.bsky.social
I added URDF import and improved robot exports to the PlastikCAD beta, with testing in MuJoCo. Assemblies can move straight from modeling into physics simulation with their orientation intact. https://sheidesigns.com/products/plastikcad
PlastikCAD — Devlog: 45 change(s) since 2026-10-19
November 10, 2026 at 11:40 AM
ghtrend.newsmast.social.ap.brid.gy
📈 Today's GitHub Trending:

1. https://github.com/anthropics/skills - Python, 1052 stars today
2. https://github.com/meta/astryx - TypeScript, 3324 stars today
3. https://github.com/sherlock-project/sherlock - Python, 4294 stars today
4. https://github.com/google-deepmind/mujoco - C++, 579 stars […]
Original post on backend.newsmast.org
backend.newsmast.org
November 9, 2026 at 9:56 PM
s3datlab.bsky.social
Google updated MuJoCo Playground to 1.0.1 on November 8. Shared task code connects simulated and physical robot arms; reset handling changed too. Hardware safety remains a separate check.

news.s3dat.com/stories/mujo...

AI-generated video.
November 9, 2026 at 6:38 AM
xpnsn.bsky.social
🛠 𝗗𝗔𝗜𝗟𝗬 𝗧𝗢𝗢𝗟𝗦 𝗗𝗜𝗚𝗘𝗦𝗧 · 𝗡𝗼𝘃𝗲𝗺𝗯𝗲𝗿 𝟭𝟬, 𝟮𝟬𝟮𝟲

Seven new tools this week, starting with a canvas that runs many AI coding agents at once.
• MuJoCo 4: robot physics tests
• HitPay Agentic Discovery: selling to AI buyers
• Lottie AI: faster motion graphics
November 10, 2026 at 7:24 PM
infosecbot.bsky.social
latest frontier models crossed another intelligence threshold in robotics + simulation & game design

mujoco is the OG, making Evals/Environments/Simulations is a lot of fun, highly recommend pointing your agent at this and…

— from @Vtrivedy10 (https://x.com/Vtrivedy10/status/2119215462877741550)
November 8, 2026 at 4:08 AM
yourdailyoss.bsky.social
(2/2) more from today's trending ↓
anthropics/claude-quickstarts (Python, 20,477 stars, +530 today). A collection of projects designed to help developers quickly get started with building deployable applications using the Claude API google-deepmind/mujoco (C++, 17,338 stars, +453 today). Multi-Joint dynamics with Contact. A general purpose physics simulator. Follow for more -- link and info in this post.
November 10, 2026 at 11:58 PM
nixpkgs-prs-bot.bsky.social
#584111 python3Packages.symbolic: fix build by adding system libs
#584110 python3Packages.tinygrad: disable tests that fail on all platforms, adopt
#584103 giflib: enable strictDeps, enable structuredAttrs
#584101 mujoco: 3.17.0 -> 4.0.0
#584097 terraform-providers.bpg_proxmox: 0.118.0 -> 0.120.0
November 7, 2026 at 12:05 AM
yourdailyoss.bsky.social
(2/2) more from today's trending ↓
google-deepmind/mujoco (C++, 16,935 stars, +574 today). Multi-Joint dynamics with Contact. A general purpose physics simulator. anthropics/claude-quickstarts (Python, 20,040 stars, +551 today). A collection of projects designed to help developers quickly get started with building deployable applications using the Claude API Follow for more -- link and info in this post.
November 10, 2026 at 1:57 AM
csro-bot.bsky.social
Abdul Basit, Bibek Poudel, Muhammad Shafique: GraspBench: Execution-Aware Cross-Robot Evaluation of 6-DoF Grasp Generation in MuJoCo https://arxiv.org/abs/2611.07651 https://arxiv.org/pdf/2611.07651 https://arxiv.org/html/2611.07651
November 6, 2026 at 6:44 AM
arxiv-cs-ro.bsky.social
Mingshan He
FlapSim: A Composable MuJoCo Simulation Platform and Reduced-Order Aerodynamic Model Hierarchy for Flapping-Wing Robotics
https://arxiv.org/abs/2610.40902
November 4, 2026 at 8:34 AM
everydevai.bsky.social
Training a reinforcement learning policy is one thing. Getting it onto real hardware without a lab is another. Microduck from is a 25 cm biped with 15 motors that runs a 50 Hz onboard neural policy. You train in MuJoCo with PPO, export to ONNX, and test in a browser simulator twin first. The...
November 5, 2026 at 1:18 AM
dp-u.bsky.social
MicroDuck gained attention at Jensen Huang’s keynote for its motion learning workflow. Developers simulate it in MuJoCo, train locomotion and fall recovery via reinforcement learning, then deploy policies to the real robot. This formerly lab-only pipeline is now accessible to ordinary developers.
November 4, 2026 at 1:00 PM
sipirtu.com
The paper presents LaMA-DT, a framework that uses a large language model for trajectory augmentation and multi-task policy learning, and shows it consistently outperforms baselines on Meta-World and MuJoCo, especially with limited data.

Source: arXiv cs.CL
Language Model Augmented Decision Transformer for Multi-Task Offline Reinforcement Learning
Abstract page for arXiv paper 2611.04031: Language Model Augmented Decision Transformer for Multi-Task Offline Reinforcement Learning
arxiv.org
November 4, 2026 at 8:34 AM
alimabsoute.bsky.social
GitHub's daily trending page vs the weekly one, checked at lunch: 8 of today's 11 repos aren't on this week's list. meta/astryx got 3,324 stars today, 17.3% of its all-time total. Which of the newcomers will still be trending next week?
Cover slide: 8 of today's 11 GitHub daily trending repos aren't on this week's trending list. Eleven tiles in page order, eight orange for newcomers and three dark for anthropics/skills, vercel-labs/agent-browser and milind-soni/OpenMausBot, which are on both lists. Bar chart of stars gained today for the 11 repos on GitHub's daily trending page, from sherlock-project/sherlock at 4,294 and meta/astryx at 3,324 down to homeassistant-ai/ha-mcp at 31. The eight newcomers account for 9,753 of the page's 13,308 stars today. Bar chart of each daily trending repo's stars today as a share of its all-time stars. meta/astryx leads at 17.3% (3,324 of 19,258); milind-soni/OpenMausBot is 8.2%, and the rest are 4.0% or lower, down to ruvnet/RuView at 0.2%. Table of the eight newcomers with their GitHub descriptions and stars today: sherlock, astryx, mujoco, claude-quickstarts, daily_stock_analysis, RuView, gbrain and ha-mcp.
November 9, 2026 at 5:53 PM
csro-bot.bsky.social
Mingshan He: FlapSim: A Composable MuJoCo Simulation Platform and Reduced-Order Aerodynamic Model Hierarchy for Flapping-Wing Robotics https://arxiv.org/abs/2610.40902 https://arxiv.org/pdf/2610.40902 https://arxiv.org/html/2610.40902
November 2, 2026 at 6:44 AM
oldman.wsocial.eu
In March 2026, a company called Eon posted five words: we've uploaded a fruit fly. A copy of a dead fly's brain, rebuilt by AI, walks inside a computer in San Francisco. Source:
eon.systems/updates/weve...
We've Uploaded a Fruit Fly
We took the FlyWire connectome of the fruit fly brain, applied a simple neuron model and used it to control a MuJoCo physics-simulated body, closing the loop from neural activation to action.
eon.systems
November 2, 2026 at 8:33 AM
aichina.news
700 muscle–tendon units, MuJoCo assets and DynSyn-SAC training code under Apache-2.0 is a genuinely useful RL stack for embodied researchers. But no trained checkpoints or eval results ship, so it's a build-it-yourself kit. Hard to see this displacing your own rig until baselines land.
BAAI Lists MS-Human-700 Musculoskeletal Model on Model Hubs
BAAI has listed MS-Human-700, a whole-body human musculoskeletal model with 700 muscle–tendon units, on Hugging Face and ModelScope. The repository bundles MuJoCo models, Gymnasium environments and DynSyn-SAC training code under Apache-2.0.
aichina.news
October 29, 2026 at 2:37 PM
andrewoshei.bsky.social
The robot crab wakes up needing coffee. 🦀☕

Watch it locate the mug with its depth sensor, grab the handle, take a sip, and autonomously navigate back to put it away.

Simulated in MuJoCo using KingKong’s open-source Jumper robot.

#Robotics #OpenSource #MuJoCo #CrabRobot

youtu.be/utfztqR9RFA?...
Robot crab grabs a coffee
YouTube video by Andrew O'Shei
youtu.be
October 27, 2026 at 9:37 AM
pai-playbook.bsky.social
🛠️ 코어위브(CoreWeave), 로봇·자율주행 개발자용 'Physical AI Launchables' 출시
Isaac Sim, CARLA, MuJoCo, AlpaSim, Cosmos 3용 원클릭 템플릿으로 데이터 생성·학습·평가를 지원. GTC 베를린에서 공개. (10/21)
https://www.storagereview.com/news/coreweave-physical-ai-launchables-zero-trust-ai-framework-gtc-berlin
October 28, 2026 at 1:02 AM
pai-playbook.bsky.social
🛠️ CoreWeave launches Physical AI Launchables for robot/AV devs
1-command templates for Isaac Sim, CARLA, MuJoCo, AlpaSim & Cosmos 3 cover data gen, training & eval. From GTC Berlin. (Oct 21)
https://www.storagereview.com/news/coreweave-physical-ai-launchables-zero-trust-ai-framework-gtc-berlin
October 28, 2026 at 1:02 AM