Subnets / SN44
Score SN44
Score (subnet 44) is a Bittensor subnet held by 6,767 unique coldkeys. Its liquidity pool holds 66.8k τ, with the alpha token trading at 0.0372 τ. Roughly 58% of tracked holders are in profit. Over the last 7 days the alpha price is down 10.2% in TAO terms. Full insider flow, whale tracking and the complete proprietary factor set for Score are available to SubnetStats subscribers.
ALPHA PRICE
0.0372 τ
IN TAO
HOLDERS
6,767
UNIQUE COLDKEYS
POOL DEPTH
66.8k τ
TAO IN POOL
7D MOMENTUM
-10.2%
ALPHA vs TAO
Project activity
that's the tweet https://t.co/260kZLwh4d
Annotation made easy.
Draw once. Apply to similar.
Sept 16. https://t.co/28qQ1P9ZGp
Score Studio is about to go live, and there’s more coming behind it.
Studio gives users and clients a place to actually use SN44 models.
That demand compounds on the subnet. Miners keep shipping SoTA expert vision models, now with a clearer path to the people who need them.
N
Gm
The GPT-6 Astra and Higgsfield sports demos deserve scrutiny before anyone calls sports computer vision solved.
That claim needs full-match accuracy measured against ground truth, alongside the cost of producing those results.
Our analysis models a dense annotation workflow at
https://t.co/ca41nguRkI
Starting next week, we’ll be inviting our clients and partners to host office hours and share videos and tutorials, on our bittensor discord channel, to help miners get better results on our private tracks.
Our current tracks are already disrupting data analysis in lower-league
And we would still be more accurate
Astra Ultra did not cook sports-grade vision AI.
Gave it a 30s football clip from our subnet private track.
Frame-level events, JSON, annotated video. Ground truth and the published scoring rules only after it committed.
22 predictions. 17 real events.
15 inside the action wi
Who’s gonna protect you when robots go full terminator?
Soon, you'll be able to build your own anti-Skynet segmentation model on Score Studio. https://t.co/QHTrKuYox6
Sunday evening reminder.
Private tracks are the part of a subnet nobody can usually check.
Ours you can.
Random challenge. Top miner response. Ground truth.
Open the console and verify it yourself.
Low League Football
https://t.co/qxN2GR9AMy
Cricket https://t.co/Z0rMaODmKF
Score Studio won't be limited to the models trained (past, present or future) on the subnet. You'll be able to run the rest of the computer vision stack in the same app.
This is our quantized SAM3, used inside Score Studio, on Djokovic vs Nadal from the 2013 US Open final.
Both
Score Studio won't be limited to the models trained (past, present of future) on the subnet. You'll be able to run the rest of the computer vision stack in the same app.
This is our quantized SAM3, used inside Score Studio, on Djokovic vs Nadal from the 2013 US Open final.
Both
A look back at how Score evolved, the decisions that shaped our direction and the roadmap ahead.
None of it was planned this neatly, but that is kind of the point.
Our ability to keep adapting is what will help us bring open vision AI to the world.
https://t.co/X3bEpNLwET
PSA: Score Studio launches on Sep 16
John Wick scenes are one of the hardest to track. we ran our subnet vision models on one of them.
Shot at night, barely any light. The camera whipping around, cars sliding sideways, people cutting through the gaps, all of it smeared with motion blur. this is the footage most mod
With all due respect, the only person we won't ever be able to detect is @JohnCena https://t.co/Tm4TxlKV4i
Grokin’
The newest GTA 6 gameplay trailer dropped yesterday. we ran our subnet vision models on it.
Person, car, gun, segmented and tracked across every scene. night, water, motion blur, freefall over the city. footage the model had never seen.
The part that matters is that this isn't
We’ve onboarded several cracked computer vision engineers as alpha testers for Score Studio.
One of them sent this video back with this note:
“Crossed arms are basically an adversarial attack and it still called it. person 0.99, dog 0.98, 42ms / if it can catch a dog mid-dance,
When we started Score, the standard computer vision tools already existed. About a million people use them every day. Most of those people are still waiting on labels, running training jobs by hand, and watching models fail once they leave the test set.
Most of those people are
Put Computer Vision on Autopilot.
Generate data. Label it. Train models. Evaluate every candidate. Deploy the winner.
Automatically.
Score Studio waitlist is open → https://t.co/RiWhQPvPZ0 https://t.co/mYmzl6BJPb
Next growth move for Score Studio:
Native Vision AI plugins for every vibe coding tool.
Starting with @cursor_ai.
Hoping to land in the official marketplace soon. https://t.co/CmiE4MpUla
Is she going to score the point?
Is volleyball going to be the next sport running on the private track?
Could Vision AI be funny sometimes?
Lots of questions and it’s just Monday… https://t.co/qMq1tIqZFA
SELLERS DRAINED
65%
28D SELLERS EMPTIED OUT
CAPITULATION
-0.68
BOUNCE COMPOSITE (z)
HEAT 7D
0.31×
VOLUME ÷ POOL DEPTH
HOLDERS IN PROFIT
58%
TRACKED WALLETS · ENTRY BELOW PRICE
See who's buying and selling Score — every trade classified miner / validator / owner / outsider, plus holder cost basis and the full 23-factor screener.
OPEN THE INSIDER TAPE →