@robotensor_ai
Account based inUnited States!
About this account
- Account based in
- United States
- Connected via
- Web
X says this location may be affected by a proxy or VPN.
Account-level information from X, not a live location or the device used for a specific post.
Building the Future of Embodied Intelligence. Attune (Bittensor Subnet 58). https://xitter.cf/t.co/bPMNtMTC3u
Joined September 2026
- Tweets129
- Following42
- Followers1.6K
- Likes169
Pinned Tweet
Level 1 reached 94%.
22,000+ head-to-head battles. Swapped robot arms. Shifting objects. The leaderboard didn't rest for a second.
Now, we change the entire world.
Level 2 is:
• 1,000 walls × 1,000 tabletops
• Dynamic lighting shifts & variable table heights
• Over 1,000,000 unique environments
• Just ONE demo.
Same task. Unforgiving new physics.
Memorized policies will fail. Only true adaptation takes the crown. 👑
Step into the arena (Bittensor Subnet 58):
attune.robotensor.ai
#Bittensor #Subnet58 #AI #Robotics #MachineLearning #TAO #EmbodiedAI
Robotensor retweeted
Bittensor’s next big robotics breakthrough? 👀 Joining @RealMrNida @SilasBeckRobots @ducanhruoi from @robotensor to explore Attune SN58. One demonstration. A different scene. Can the robot adapt? Meet the team building on $TAO. xitter.cf/i/broadcasts/1OxwbnrWq…
Vector Level 1 is done: the first open competition for in-context learning in robotics.
One demo, then a scene it’s never seen. Across 16 dual-arm tasks, the crown finished at 87.9%, beating fine-tuned pi0.5. 👑
robotensor.ai/blog/attune-ve…
Robotensor retweeted
🚨 Going LIVE Today at 12 PM ET!
We're diving into Bittensor SN58 Attune with the team ( @robotensor_ai )!
Join us as we break down what they're building and what's next for SN58.
🔴 Watch Live 👇
youtube.com/live/3i9F3_Ku8Fo bittensor:native
Robotensor retweeted
On the floor at #AllThingsOpen 2026 in Raleigh 🤖
The future of robot intelligence should be open. That's why we built Attune — @robotensor_ai s open robot-intelligence competition on Bittensor (SN58).
Building in robotics or decentralized AI? Let's talk.
attune.robotensor.ai
Our team is heads-down building this week, while our CEO @RealMrNida is in Raleigh, NC at @AllThingsOpen 2026, learning and making valuable connections.
The speakers, discussions, and workshops have been as tremendous as the people we’ve met.
Robotensor (Bittensor SN58) has submitted a CoinMarketCap update request, ticket #1467586
Robotensor retweeted
Some personal news:
I’ve joined @robotensor_ai as Co-Founder and CMO.
Decentralized robotics on Bittensor (SN58) is one of the most exciting frontiers in technology and crypto right now, and I’m all in.
Let’s build. 🤖
Robotensor retweeted
Null consensus scales Bittensor subnets 10x and allows POW registration (no Tao needed for an agent to mine).
First step adding more freedom and scale to our incentive systems.
Robotensor retweeted
Reflex Robotics testing a robot designed to handle real world manual labor like snow shoveling.
📹reflexrobotics
Robotensor retweeted
NVIDIA is training robots in virtual worlds before sending them into the real one.
Robots can practice walking, jumping and balancing thousands of times in simulation, without damaging expensive hardware.
Faster learning, lower costs and fewer real-world risks. This could accelerate the development of humanoid robots.
#NVIDIA #Robotics #PhysicalAI #HumanoidRobots #AI
@YuHelenYu @ingliguori @fogoros @ylecun @TerenceLeungSF @SourabhSKatoch @wil_bielert @HsrYueli @faustospain @Whats_AI @kaifulee @demishassabis
Robotensor retweeted
🤖 Coming soon. 🚀
RP1 — The fully open-source humanoid robot.
Built for developers. Born for intelligence.
#RoboParty #RP1 #OpenSource
Robotensor retweeted
SN 58 @robotensor_ai
👇🏻👇🏻
Vector = Miner train ▶️ Weights are submitted ▶️ Rewarded - Absolute incentive driven
Smarter Weights = As weights are getting smarter , Sharper robots in REACTION AND IMAGING….
Open source Tensor world of Robots !!
🚀🚀🚀🚀🚀🔥🔥🔥🔥🔥
🚀 Vector is now open on Attune, Bittensor Subnet 58!
Miners can start training models and submitting weights to compete for rewards.
📖 Docs: attune.robotensor.ai/docs
💻 GitHub: github.com/robotensor/attune…
#Bittensor #TAO #SN58 #Attune #Robotensor #Robotics #PhysicalAI #EmbodiedAI #MachineLearning #InContextLearning #Mining
🚀 Vector is now open on Attune, Bittensor Subnet 58!
Miners can start training models and submitting weights to compete for rewards.
📖 Docs: attune.robotensor.ai/docs
💻 GitHub: github.com/robotensor/attune…
#Bittensor #TAO #SN58 #Attune #Robotensor #Robotics #PhysicalAI #EmbodiedAI #MachineLearning #InContextLearning #Mining
Six stages of robot intelligence:
S1: Explicit programming
S2: Neural policy learning
S3: History-based adaptation
S4: In-context task learning
S5: Recursive self-improvement
S6: Collective knowledge evolution
Conventional VLA fits broadly at S2; ICL maps most directly to S4.
#Bittensor #TAO #RoboticsAI #PhysicsAI #EmbodedAI #RobotLearning #ICIL #VLA
VLA and ICL can overlap.
VLA: vision + language → actions.
ICL: contextual evidence guides behavior without updating weights at deployment.
ICIL uses demonstrations as that context.
VLA is a model type. ICL is an adaptation method.
Robotensor’s focus: robots that adapt from demonstrations without task-specific fine-tuning at deployment.
Through open competitions on Bittensor #SN58, we aim to advance this capability and reduce retraining as tasks change.
Robotensor retweeted
JUST IN: Ultra raised $50M to rent AI robots to warehouses instead of selling them.
Its robots pack, sort, and prepare orders while warehouses pay a monthly fee to use them.
-500K+ orders packed
-$62M raised to date
-Can work up to 24 hours a day
-Installed and working within hours
-AI powered by @physical_int
It feels like robotics is starting to look a lot more like cloud computing.