@OAlsta
Account based inUnited States!
About this account
- Account based in
- United States
- Connected via
- United States Android App
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.
Perfect Resonance in All Things 》wu wei《
Mediterranean California
Joined October 2021
- Tweets21K
- Following4.4K
- Followers791
- Likes85.4K
Pinned Tweet
☆ Creation of a 1000 year Federation ☆
▪︎ $TAO
▪︎ @bittensor
Blake de Cleyre retweeted
Beam is building in the open.
Don't trust us. Verify.
Every transfer. Every assignment. Every score. Every reward calculation.
No black boxes.
api.b1m.ai/v1/audit/transfer…{id}
Explore the full audit endpoints 👇🎬
Beam | SN105 | Bittensor
Blake de Cleyre retweeted
Using one killer feature (@Hippius_cloud screen-sharing) to demo another (doug Ai reports from @taostats).
console.hippius.com/share/Fv…
Blake de Cleyre retweeted
Spare compute, something brilliant.
.feat @macrocrux and @mccrinbc.
Blake de Cleyre retweeted
YOUR HERMES AGENT CAN HAVE ITS OWN DESKTOP. WATCH IT WORK LIVE, TAKE CONTROL, AND HAND IT RIGHT BACK.
This is one of the standout capabilities included in Hermes v0.22.0, The Threshold Release. 🔥
It's called Bot Screen, and here's what makes it so cool:
➤ Its Own Desktop
Your Hermes bot gets a dedicated desktop on a Linux gateway, where it can use browsers, terminals, and apps without taking over your computer.
➤ Watch It Work Live
See exactly what your agent is doing on its desktop, in real time.
➤ Take Control Anytime
Need to handle a login, 2FA, or CAPTCHA? Jump in with your mouse and keyboard, complete the step, and hand control right back to Hermes.
➤ Keep Working Remotely
Your bot can continue working on the gateway even after you close Hermes Desktop or shut your laptop.
Bot Screen first landed a few weeks ago, but it's worth highlighting again as part of the new stable release.
If you missed it the first time, check out this demo! 🔥
Blake de Cleyre retweeted
JUST IN: SN80 (@openroboto), the largest robotics subnet on @bittensor, integrates peaqOS
All devices now have an ID and wallet on Solana, and can be verified and paid for their performance directly on Solana using @peaq
Blake de Cleyre retweeted
Actual Computer turns your NVIDIA DGX Station into a department-scale AI factory.
Read the full case study:
actual.inc/company/blog/case…
Blake de Cleyre retweeted
Normally when you load a model with other leading engines, you're forced to wait minutes before you get a fully functional server.
When you use Actual thats not the case. Cold starts in less than 20 seconds and cached starts in less than 4 seconds!
best engine everrr
Blake de Cleyre retweeted
@actualinc is getting early access to nvidia releases, listed as inference partner on @nvidia site and is dropping banger updates every other week.
I get it that comms around mining are lacking, but is it really so bad to build out a rich demand layer before launching mining?
That seems to be what they are doing.
Too many subnets launch mining before there is really anything meaningful to mine. Or before their service actually has traction.
I could be wrong, but having such deep connections with big players like Hermes, nvidia and prime intellect is a lot of smoke.
Blake de Cleyre retweeted
We are going to make sure that the US catches up and then smokes the competition when it comes to Open Superintelligence
Nvidia’s Nemotron 3.5 Super VL is now the leading US open weight model, scoring 26 on the Artificial Analysis Intelligence Index
Nemotron 3.5 Super VL achieves twice the intelligence of its predecessor, Nemotron 3 Super (13), and does so using ~60% fewer output tokens per task
It launches alongside Nemotron 3.5 Bolt 74B-A9B 24GB, the first model in its new Bolt family, which scores 21 on the Intelligence Index
Congratulations to @NVIDIAAI on the releases!
Key Takeaways:
➤ Leading US open weight model: Nemotron 3.5 Super VL scores 26.5, edging out Thinking Machines' Inkling Small (25.7) and ahead of NVIDIA's Nemotron 3 Ultra (23) and Meta's Muse Glimmer (17)
➤ Bolt comes within 2 points of Ultra: Nemotron 3.5 Bolt 74B-A9B 24GB scores 21, 1.5 points behind Nemotron 3 Ultra 550B A55B (23). The 2 bit memory has about a seventh of Ultra's parameters (74B vs 550B) and a sixth of its active parameters (9B vs 55B)
➤ Super 3.5 VL near lossless in NVFP4: Nemotron 3.5 Super VL releases in NVFP4, scoring within 0.1 points of its BF16 weights on the Intelligence Index, with no component moving by more than 2.6 points
Nemotron 3.5 Super VL details
Context window: 1M tokens
Modalities: Vision and text input; text output
Parameters: 120B / 12B
Nemotron 3.5 Bolt 74B-A9B 24GB details
Context window: 256k tokens
Modalities: Vision and text input; text output
Parameters: 74B / 9B
Blake de Cleyre retweeted
Unicorn incoming
We've developed a new inference engine, called 'e'.
It's the closest thing to alien technology which we have, and pushes 10,500 tok/s aggregate on a single DGX Station with the new @NVIDIAAI Nemotron 3.5 Super VL model, released a few minutes ago.
Blake de Cleyre retweeted
We've developed a new inference engine, called 'e'.
It's the closest thing to alien technology which we have, and pushes 10,500 tok/s aggregate on a single DGX Station with the new @NVIDIAAI Nemotron 3.5 Super VL model, released a few minutes ago.
Actual Computer turns your NVIDIA DGX Station into a department-scale AI factory.
Read the full case study:
actual.inc/company/blog/case…
Blake de Cleyre retweeted
How does the EAQ v1 performs?
The idea is to use LOWER bit quant and still be at least AS smart as a HIGHER bit EXL3.
In Qwen3.8-27B, not only that, it BEATS it in agentic tool use by a large margin, while being 22% smaller!
- Qwen3.8-27B EAQ 3.0 BPW - 12 GB
- Qwen3.8-EXL3 3.5 BPW - 15.34 GB
This gives you an idea of what EAQ is going to be, and this is just an experimental v1!
COMING SOON !
Blake de Cleyre retweeted
The upload price: what changed?
Every agent submission creates work: screening, execution and validation.
Ridges now has a dynamic upload-pricing model for each competition.
Under the default settings:
• The minimum price is $5.
• Each purchased upload raises the price by approximately 15%.
• Between purchases, the price halves every 30 minutes, down to the $5 minimum.
Planning your next submission? Check the live upload price. A quieter period could mean a lower entry cost.
Blake de Cleyre retweeted
actual computer
turns your local computers
into a department- scale AI factory
@actualinc
Today most teams rent their agent swarms from hosted models and pay for every token.
At list prices, 32 Luna subagents running around the clock cost $4,686 a month.
With Actual, the swarm runs on Nemotron from your DGX Station instead.
Blake de Cleyre retweeted
I'm so excited to be releasing this engine.
This represents the first of hopefully a long line of work with Nvidia
We didn't put comparison charts in the main post
but rest assured
WE SMOKED EM
this is the best engine for your dgx station.
its not close.
Actual Computer turns your NVIDIA DGX Station into a department-scale AI factory.
Read the full case study:
actual.inc/company/blog/case…
