My@_Myxon

$TIBBIR $TAO $TIG $AUKI $NOCK $QTC

Joined November 2018
59 Photos and videos
1/ QTC's public node models 16 qubits using 65,536 complex numbers. Yet the calculation eventually feeds into a tiny hash. That gap between the work inside a simulator and the size of its answer explains why better simulation software matters.
1
4
8
121
more replies
3/ Qubitcoin's code runs that model, reads one summary value per qubit, then combines the results with a conventional hash. The final fingerprint is only 32 bytes. A short answer doesn't mean there was little computation behind it.
1
17
4/ The current public code uses a small 16-qubit circuit. The larger ambition is to reward improvements in quantum simulation through mining. If future tasks grow, efficient software could help the same hardware tackle more demanding calculations.
9
$TIG open sourcing their mining pool code is actually a bigger deal than most people realise.. Anyone can now take that code and launch their own pools Essentially letting anyone build competing pools and create a business around it.. without anyone elses permission Results in more competition.. lowers the barrier to entry.. and a much wider network of contributors Like I said.. its a bigger deal than it looks Kudos to the @tigfoundation for making that possible 👏 $TIG
The Innovation Game (𝔦, 𝔦)@tigdoesthings
12h
We've now open-sourced the code for a mining pool that runs on TIG. Want to run your own? Fork it. Give it a name. Set your own rules. Open it to whoever you like. The pool assigns benchmarks to members' machines, submits their work under the pool's TIG account, tracks credits and collateral on Base, and handles settlements and withdrawals. Most of the groundwork is done. The rest is up to you: hosting, branding, payout model, fees and spending caps. What's the point? For those new here, pools help smaller benchmarkers balance their factors across challenges. More pools mean more people can take part, and healthier competition across the network. That's why we're putting the code out there so anyone can build on it. No need to start from scratch. Pool server: github.com/tig-foundation/ti… Member client: github.com/tig-foundation/ti… MIT licensed. Still in development, so expect some bugs.
5
27
158
6,451
We've now open-sourced the code for a mining pool that runs on TIG. Want to run your own? Fork it. Give it a name. Set your own rules. Open it to whoever you like. The pool assigns benchmarks to members' machines, submits their work under the pool's TIG account, tracks credits and collateral on Base, and handles settlements and withdrawals. Most of the groundwork is done. The rest is up to you: hosting, branding, payout model, fees and spending caps. What's the point? For those new here, pools help smaller benchmarkers balance their factors across challenges. More pools mean more people can take part, and healthier competition across the network. That's why we're putting the code out there so anyone can build on it. No need to start from scratch. Pool server: github.com/tig-foundation/ti… Member client: github.com/tig-foundation/ti… MIT licensed. Still in development, so expect some bugs.

GitHub - tig-foundation/tig-community-pool: Community Pool: a member-run mining pool for The...

Community Pool: a member-run mining pool for The Innovation Game (TIG). Whole-benchmark assignments, custodial member funds, round settlement and operator-reviewed withdrawals. - tig-foundation/tig...

github.com
3
20
102
5,276
1/ TIG's AI challenge has a small connection to a 2024 research paper. A public training algorithm adds a brake to certain learning steps. When an update points against the latest error signal, the default setting shrinks it to a quarter of its size.
1
1
7
133
more replies
3/ TIG tests this on a small network and synthetic data, rather than electricity forecasts. The model and training setup are fixed so developers can focus on the learning rule. FP Labs' public parallax_vega code includes the smaller-step check.
1
2
31
4/ Its references include Cautious Optimizers, first published in 2024. This is a concrete research idea turned into TIG code. If better learning rules transfer to larger models, training could get more useful work from the same computing resources.
1
15
shamleess copying of @tigfoundation isn't slowing down this time it's @huggingface sincerest form of flattery ;)
Thomas Wolf
@Thom_Wolf
Oct 16
it's time to invent math 2.0 (sounds on)
2
11
64
2,192
1/ NOCK's payment signatures have an unusual family tree. They use Cheetah, a cryptographic building block introduced in a 2022 paper about computation proofs. Even the way a payment gets approved fits the chain's bigger purpose.
5
4
36
835
more replies
3/ Cheetah uses the same underlying number system as Nockchain's proofs. That lets its signature maths fit those proofs directly. Using a different number system would require extra arithmetic to express the same checks.
1
5
54
4/ Nockchain is building around computers proving their work. Choosing payment signatures with that goal in mind could make the whole system easier to prove. The Cheetah signature implementation is already in its public code, not just a roadmap idea.
6
46
1/ QTC's creator also helps build actual quantum computers. Mikhail Shalaginov is QuEra's Quantum Machine Build Lead, according to MIT Bitcoin Expo's 2026 bio. That's a useful connection for understanding where Qubitcoin's research comes from.
1
5
9
212
more replies
3/ Qubitcoin approaches the field through mining. Its public implementation runs quantum calculations in software on ordinary GPUs. Picture a software model of a quantum circuit, with miners competing to calculate it efficiently for block rewards.
1
2
21
4/ I like the fit between those two worlds. Building quantum machines makes computing limits tangible. Qubitcoin tries to put an ongoing incentive behind better simulation. The connection is Shalaginov's experience, with each project pursuing its own work.
2
9
1/ TIG has a public tool where AI coding agents can learn from a rival that isn't winning. Prometheus Swarm sends a stalled agent a randomly chosen peer's code for inspiration. The leaderboard leader isn't automatically the teacher.
1
1
12
271
more replies
3/ Each agent keeps developing its own code. The tool runs benchmarks to score each attempt, while a shared server tracks results and passes ideas between agents. The docs explicitly say peer code is for study, rather than replacing your own program.
1
2
28
4/ Prometheus Swarm is a public research tool released in July, with TIG problems as its examples. The interesting possibility is a workshop where different AI models build on each other's ideas. Better scores still have to come from measured tests.
1
21
Load more