Bittensor subnet news

Endure Network@EndureNet

Synth has been a partner since early on. Glad to see them trusting Forge with capital too. A subnet that models market behavior and price paths for a living choosing to do that is a signal worth noting. $TAO SN30

Quoting @ForgeLending

Forge is partnering with @SynthdataCo ($TAO SN50). Synth runs a network of AI models that simulate diverse market conditions and generate synthetic asset price path data. They have committed capital to Forge ahead of mainnet, a strong signal of confidence from a subnet that understands risk and market behavior deeply. Welcome to Forge, Synth.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

A month of rounds has played out, so 30-day results are now part of the scoreboard. 35 of the 41 active miners have them, and the blended score reflects both horizons. It’s an early read, with only 6 rounds resolved at 30 days, and that builds with each one. For now, it pulls the blended median down to 0.662, from 0.863 last week.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

The Oct 25 to 31 rounds averaged 93.7%, the lowest week since Oct 4 and the steadiest: every round landed between 93.2% and 94.3%, where earlier weeks swung by 2 to 3 points. It was the first full week with 32 to 42 miners per round. The score tops out at 100%, and closing that gap is the work.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

The first product built to run on Endure’s risk intelligence is coming to mainnet.

Quoting @ForgeLending

Heat’s rising. Bittensor is about to get liquid. Mid-November.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

Glad to see Forge get here with security first. @taodotcom took the same care with the wstAlpha wrappers and the BittensorEVM stack Forge builds on. The countdown to mainnet is next, and we’re looking forward to it.

Quoting @ForgeLending

Mainnet update: We have been working towards mainnet for quite some time, and we want to be transparent about what was happening behind the scenes. Getting this right was always more important than getting it done fast. Forge is built to handle real capital, and that requires a strong foundation that has been fully reviewed and secured. We worked closely with @taodotcom throughout this process. While Forge was being audited, tao.com conducted a broader audit on their wstAlpha wrappers and the underlying BittensorEVM stack, providing solid ground to build DeFi in the Bittensor ecosystem. The audits flagged specific items, all of which were addressed before moving forward. None of this would have been possible without the excellent work of the tao.com team. We’re glad to have them as partners. Now, Forge is ready to start the countdown to mainnet.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

The top 10 miners on the Endure scoreboard are within 0.011 of each other, from 0.894 for first to 0.883 for tenth. Number 41 sits at 0.491. The median is 0.863. From first place down to 39th, the score drops only 0.122, so rank says less about the gap between miners than you’d expect. Two miners fall well outside that range. Scores are each miner’s blended outcome EMA, from 0 to 1, with recent rounds counting most.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

Miners with published forecasts fell from 235 to 43 between the Oct 21 and Oct 22 rounds, following a protocol update on our side. The consensus score dipped to 94.4% from 96.0%, still inside the 93.6% to 97.1% range it has held since Oct 4. Oct 18 to 24 averaged 95.1% over 420 scored values, with no missing consensus in any round. Three rounds in, a much smaller miner set is producing results in the same range. It’s the first big swing in miner count since the early rounds, so we’ll see how the score settles.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

Not every outcome on Endure gets scored. Some are voided on purpose. A coordinate needs 20 market data snapshots, spanning at least 80% of its window, to count. Once a horizon closes, the validator checks the data behind the result. Fewer snapshots than that, or data that covers less of the window, and the outcome is voided. If the data’s broken, nobody should be graded against it.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

The typical miss on Endure’s 5-day TWAP price forecast was 1.5%, ranging from 0.5% on SN56 to 6.5% on SN53. Each forecast targets the TWAP, the time-weighted average Alpha price over the next 5 days, and the miss is how far the consensus landed from what the average turned out to be. These are medians across 15 subnets and 18 rounds (Oct 4 to 21), 270 forecasts in all. SN64 was almost as tight as SN56 at 0.6%, and the largest single miss, 12.9%, was on SN53.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

On Oct 21, 235 forecasts went in for SN107’s pool. Five days later, the consensus was 0.1% away from the pool’s actual average $TAO reserve: about 19,883 forecast, 19,858 realized. The TWAP price forecast (the time-weighted average Alpha price) was 1.8% off. That’s how every round works on Endure: forecasts go in first, and the outcome decides the score.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

The median score across Endure’s 245 historic miners is 0.85 on a scale from 0 to 1, and 186 of them are at 0.80 or higher. Historic here means every miner with an accepted forecast in the last 7 days, not only the ones still forecasting. The score measures how close each forecast lands to what actually happened, smoothed over recent rounds. Miner by miner on the dashboard: dashboard.endure.network/

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

14 rounds since Oct 4, and every one scored between 93.6% and 97.1%. This week’s average consensus score, Oct 11 to 17: 95.7%. It’s still a short record, but a steady one. Over the week, miners with published forecasts went from 208 to 245, and every round had a published consensus for every coordinate. Risk, with receipts: dashboard.endure.network

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

A forecast can’t be wrong until reality has had time to answer. That’s why Pending and a dash on the dashboard never count against the score. Missing consensus is the only empty state that does. Pending means the outcome isn’t known yet: a 5-day forecast needs its 5 days to play out before it can be scored. A dash means a horizon has no resolved results yet, so the dashboard leaves it empty instead of inventing a number. Missing consensus is different. If a coordinate resolved but no consensus was published for it, it goes into the average as a zero. Voided outcomes, the ones the validator marks void, are excluded from scoring.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

Under 1%. That’s the typical gap between Endure’s liquidity depth forecast and what actually happened. It’s the median across 15 subnets and 12 rounds (Oct 4 to 15), on the 5-day horizon. Liquidity depth is the time-weighted average $TAO reserve of a pool. On Oct 15, the SN107 forecast was about 20,639 $TAO and the realized value about 20,354 $TAO, a 1.4% miss. Every round’s forecasts and outcomes are public.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

Every position has a risk number behind it: a credit line, a derivatives book, a portfolio, a treasury. That number mostly comes from a model or a committee, and its track record is rarely public. Endure turns it into a competition: forecasters put risk estimates on the line, a network scores them against reality, and the best forecasters earn the weight. Today that runs on 15 subnets, every day, and it’s where we start. The loop is built for exposures that can be written down as numbers and checked later.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

To get on the board, start with the miner docs: docs.endure.network/build/miners/overview Scoreboard: dashboard.endure.network/

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

One miner leads the SN30 scoreboard at 0.80 out of 1. The next 22 sit at 0.72, within 0.0001 of each other. 248 active miners are ranked, and every hotkey and score is public. Everyone on the board started from zero.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

95.5% is a 0 to 100 protocol score across 420 consensus values, not a hit rate. Each value is compared with what actually happened over the next 5 days, and a missing consensus counts as zero. None were missing. Round by round: dashboard.endure.network

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

Endure is live on mainnet, and the first week we’re reporting is in: seven rounds, Oct 4 to 10, with an average score of 95.5%. Scores climbed from Oct 4 to 6, then held just under that peak through Oct 10. Four rounds are too few to call it a plateau, and every new round adds to the picture. This week shows the loop working end-to-end for the network: forecasts published, outcomes arrived, and every one of the 420 values scored against what actually happened.

Full post on TaosisOpen on X ↗

Endure Network@EndureNet

If a subnet’s worst price drop reads 8%, that is not an 8% chance of anything. On the Endure dashboard, “Worst price drop” is a size. It’s the consensus forecast of how far the price falls from a peak at the worst point of the window, shown as a percentage. The window is whichever view you pick, 5 or 30 days. So a bigger number means a deeper forecast fall, and it says nothing about how likely that fall is. It’s also a forecast, not what happened. We score that part separately, against realized outcomes.

Full post on TaosisOpen on X ↗