Introducing Zils, Subnet 127 on Bittensor.
We train custom System One models for your specific use case. Upload your examples, miners compete to train the best model, validators grade it, you get an API.
Sort, flag, score, act. Small models that are right for your data.
https://t.co/5ufRzx3CpW
AI agents need more than autonomy. They need evidence at every handoff, with people retaining control.
At @AIBC_World Rome, @willobrien joined @macrozack of @bitstarterAI to discuss governed autonomy and open access to compute, operations and capital.
#AI#Bittensor#SN87
We've had a lot to share with the Bittensor community and plenty more to come.
On top of our own deep thinking about how to apply the Carbon model to Bittensor, in the lead up to @ExploitSummit, the @bitstarterAI team pushed us to get really clear on the what, how & why.
Thanks for talking shop with @RyanBequette@mccrinbc & @macrozack
@Pop_Collapse Yes, teams directing value back into their SN outperform those that don't - but not all businesses are in a position to do so. Not sure about this specific case, though
Saw your question about using agentic trading to fund alpha buybacks - we're setting that up on @cascade_sn91
Wait, Synth's miners outperform Opus 5.5 - and GPT-6 Google Gemini
Against a live benchmark for agentic AI quant activity, over 30 days Synth won easily
Am I reading this right bittensor:native
A live benchmark for AI-driven quant research.
Challengers: @AnthropicAIโs Opus 5.5, @Googleโs Gemini and @OpenAIโs GPT-6 Astra.
One briefly hit #1. None sustained a win over Synthโs best quants over 30 days.
Results, code, methodology:
https://t.co/r3z6nnQgzq
๐จ Subnet Summer TAO X Spaces - October 28, 2026 | Wednesday at 5PM GMT
๐ง๐ผ๐ฝ๐ถ๐ฐ: ๐ง๐ต๐ฒ ๐๐ถ๐ฟ๐๐ ๐ฃ๐ผ๐๐-๐๐ฎ๐น๐๐ถ๐ป๐ด ๐ฌ๐ฒ๐ฎ๐ฟ: ๐ช๐ต๐ฎ๐ ๐๐ ๐ฅ๐ฒ๐๐ฒ๐ฎ๐น๐ฒ๐ฑ ๐๐ฏ๐ผ๐๐ ๐ง๐๐ขโ๐ ๐ฃ๐ฎ๐๐ต ๐๐ผ ๐ฎ๐ฌ๐ฎ๐ณ
The halving didnโt just cut emissions. It stress-tested every part of Bittensor at once.
Four voices, four angles: capital, on-chain data, building, and validating.
Speakers:
โข Mark Jeffrey (@markjeffrey) - @stillcorecap
โข Mog (@mogmachine) - @taostats | @hippius_cloud | @vidaio_ | @blockmachine_io | @say_gm_
โข Max Sebti (@MaxSebti) - @webuildscore (SN44)
โข Tribit (@iamtheloraxnow) - Medulla Labs | @subnetsummerTAO
โข Bryden (@SubnetSummerT) - @subnetsummerTAO
Join us this Wednesday for a high-signal conversation on what the first post-halving year means for Bittensorโs path to 2027.
Set your reminder and join the discussion.
100% - if you havenโt come across @tensorlink_dev or his work on @cascade_sn91, youโre missing one of Bittensorโs best builders.
Post below explains how you can deploy our time series API as forecasting infrastructure, โsaving weeks of guesswork.โ
The Bittensor community is as strong as ever and gets stronger by the day.
Along my journey I've found many great people and @tensorlink_dev from @cascade_sn91 is one of those great people.
Massive thanks to Chris he has saved us weeks of guesswork.
We run Cascade's synthetic time-series generator as a research substrate. The Alphacast feed sits alongside it as forecasting infrastructure. Both have become core pieces of our engineering stack.
Research-stage, and one live deployment which we constantly update,the capability jump has been huge. I can't go into to much detail but I can say this,without the Cascade tech we wouldn't really have an edge. Not only is the tech been built on SN91 brilliant it's highly versatile. It's like giving an acromat artist color vision which really broadens your horizons Cascade is a truly remarkable technology.
gScade๐ซก
Get the MCP: https://t.co/g4EKNWj8Dy
Join the Telegram: https://t.co/XJc4TLHzmE
โYouโre not launching a company on Bittensor, youโre forming a network.โ
Your miners will challenge your design choices. Theyโll try to break what you build. Founders need to be ready for that.
At Exploit Summitโs incubator panel, Bitstarter co-founder @macrozack explained what founders need to get right: what miners do, what validators score, and what the subnet produces.
If you canโt explain that clearly, how can anyone back what youโre building?
Watch Chrisโs take below.
From 4M to 1B+. Cascadeโs biggest upgrade yet.
Weโre preparing to train Cascade-MoE: a mixture-of-experts time-series foundation model targeting 1B+ total parameters.
Weโve shown that synthetic data generated by Bittensor miners can train a 4M-parameter forecasting model from scratch, reaching parity with Toto 2.0 4M. Now weโre scaling up, with SOTA forecasting as our goal.
Cascade-MoE will use a learned router to select combinations of expert networks based on patterns in the input, alongside an always-active shared expert. This gives the model access to specialised capacity without activating every stored weight for every input.
Weโre also evolving the competition so miners can generate synthetic data targeted at particular domains and time resolutions, such as hourly versus daily series.
The architecture is designed to support targeted training of selected expert subsets, while the subnet tests whether contributions improve forecasting performance.
The goal: SOTA forecasting accuracy, built through competition on #SN91.
SN91 is now training a 1B parameter model.
If 1B sounds small: the new model will rank among the largest time series foundation models, ever.
And our current model - with only 4M params - already beats Salesforce's 91M model.
Which means: we're going after SOTA.
From 4M to 1B+. Cascadeโs biggest upgrade yet.
Weโre preparing to train Cascade-MoE: a mixture-of-experts time-series foundation model targeting 1B+ total parameters.
Weโve shown that synthetic data generated by Bittensor miners can train a 4M-parameter forecasting model from scratch, reaching parity with Toto 2.0 4M. Now weโre scaling up, with SOTA forecasting as our goal.
Cascade-MoE will use a learned router to select combinations of expert networks based on patterns in the input, alongside an always-active shared expert. This gives the model access to specialised capacity without activating every stored weight for every input.
Weโre also evolving the competition so miners can generate synthetic data targeted at particular domains and time resolutions, such as hourly versus daily series.
The architecture is designed to support targeted training of selected expert subsets, while the subnet tests whether contributions improve forecasting performance.
The goal: SOTA forecasting accuracy, built through competition on #SN91.
We provide the full-stack for time-series foundation models.
Training, fine-tuning, benchmarking, inference, harness-creation.
We build models on #Bittensor, offer SOTA TSFMs through Ephemeris, and create tools to connect your time-series data to any agent of your choice.
See our current model: https://t.co/Si8FoWHe6t
Get Ephemeris inference via MCP: https://t.co/wVutnqEDoZ
P R E S S S T A R T โถ๏ธ
New teams. All stars. Pure alpha.
Let's get #Bittensor kickstarted ๐
B I T S T A R T E R . . . L O A D I N G ๐ #bitstarter
We just killed cold outreach.
Introducing Leadpoet: the AI sales agent that finds buyers who need you now, writes personalized outreach, and follows up on LinkedIn and email.
Used by Dropbox, OnePatch (YC26), Muttdata.
Comment โPoetโ for 100 free lead credits to find qualified buyers.