robofluent commited on
Commit
a5cde25
路
verified 路
1 Parent(s): bc8dfac

Bring the organization card up to date with Attune

Browse files
Files changed (1) hide show
  1. README.md +25 -9
README.md CHANGED
@@ -9,27 +9,43 @@ pinned: false
9
 
10
  <p align="center">
11
  <a href="https://www.robotensor.ai">
12
- <img src="https://huggingface.co/spaces/robotensor/README/resolve/main/assets/robotensor-x.jpg" alt="Robotensor logo from @robotensor_ai on X" width="112" height="112" />
13
  </a>
14
  </p>
15
 
16
  <h1 align="center">Robotensor</h1>
17
  <p align="center"><strong>The intelligence layer for robots.</strong></p>
18
- <p align="center">Show the task. Let the robot learn in context.</p>
19
 
20
  <p align="center">
21
  <a href="https://www.robotensor.ai">Website</a> 路
 
22
  <a href="https://github.com/robotensor">GitHub</a> 路
23
  <a href="https://x.com/robotensor_ai">X</a> 路
24
- <a href="https://competition.robotensor.ai">Competition</a>
 
25
  </p>
26
 
27
- Robotensor is a physical AI company building policies for dexterous manipulation.
28
- We work on turning demonstrations, perception, and action history into precise
29
- robot behavior, with a focus on in-context robot learning: adapting behavior from
30
- experience while a model's weights stay frozen.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
31
 
32
  ---
33
 
34
- Explore [open roles](https://www.robotensor.ai/careers), or
35
- [get in touch](mailto:hello@robotensor.ai).
 
9
 
10
  <p align="center">
11
  <a href="https://www.robotensor.ai">
12
+ <img src="https://huggingface.co/spaces/robotensor/README/resolve/main/assets/robotensor-x.jpg" alt="Robotensor" width="112" height="112" />
13
  </a>
14
  </p>
15
 
16
  <h1 align="center">Robotensor</h1>
17
  <p align="center"><strong>The intelligence layer for robots.</strong></p>
18
+ <p align="center">Teach robots new work by showing them.</p>
19
 
20
  <p align="center">
21
  <a href="https://www.robotensor.ai">Website</a> 路
22
+ <a href="https://attune.robotensor.ai">Attune</a> 路
23
  <a href="https://github.com/robotensor">GitHub</a> 路
24
  <a href="https://x.com/robotensor_ai">X</a> 路
25
+ <a href="https://www.linkedin.com/company/robotensor">LinkedIn</a> 路
26
+ <a href="https://discord.gg/An3YtsAaa6">Discord</a>
27
  </p>
28
 
29
+ Robotensor is a Physical AI research and development company. We build frontier robot
30
+ intelligence that learns new work from demonstrations, adapts as the work changes, and
31
+ carries across robot platforms. Our focus is in-context robot learning: a model that picks
32
+ up a new task from what it is shown, with its weights frozen.
33
+
34
+ ## Attune
35
+
36
+ [Attune](https://attune.robotensor.ai) is our open competition for adaptive robot
37
+ intelligence, run as Bittensor subnet 58. Miners train robot models; validators score them
38
+ in simulation under the same conditions, and every result is published in full.
39
+
40
+ | Competition | Focus | Status |
41
+ |---|---|---|
42
+ | [Vector](https://attune.robotensor.ai/vector) | Action-aware in-context learning: learn a behavior from one demonstration, adapt the execution | Live |
43
+ | [Horizon](https://attune.robotensor.ai/horizon) | Human-video in-context learning with World Action Models | Coming soon |
44
+
45
+ - **Mine:** `pip install robotensor-attune`, then read [what to submit](https://attune.robotensor.ai/docs/vector/rules/what-you-submit).
46
+ - **Validate, or read the code:** [attune-subnet](https://github.com/robotensor/attune-subnet)
47
+ - **More:** [Docs](https://attune.robotensor.ai/docs) 路 [White paper](https://attune.robotensor.ai/whitepaper.pdf) 路 [Taostats](https://taostats.io/subnets/58) 路 [Attune on Discord](https://discord.com/channels/799672011265015819/1550516268002578432)
48
 
49
  ---
50
 
51
+ Explore [open roles](https://www.robotensor.ai/careers), or [get in touch](mailto:hello@robotensor.ai).