NEB 🧠 The first 263 blocks of the testnet.
Each frame is one block. The neurons are the real weights of the chain's model, recomputed from the shares miners found, and a real digit goes through the network at every step.
Accuracy went from 12.7 % to 88.4 %.
https://t.co/MDp4xvWrYH
@fin_yac NEB's models are free and public. The TESS labels come from Astronet-Triage-v2, it's in the docs.
Out of 2,000 stars Kepler recorded but never targeted, the Kepler model kept one signal. A second team recovered it. It's still a candidate.
https://t.co/i4GiGpHabH
NEB miners trained a Kepler classifier that kept one planet candidate: KIC 6691030, dimming every 4.77 days. A NASA Goddard team recovered the signal: a transit fits the dip better than a sine wave. It's a candidate, not a confirmed planet. Via TraSHDUMP https://t.co/MxlZN2CaUP
We are extremely pleased to list #NeuralEngineByte!
Where miners train a neural network.
$NEB
@silas_eigen
https://t.co/vMjC0unZkD
https://t.co/06SnZtKNEu
A Kepler classifier trained by NEB miners kept one transit candidate in the Kepler Bonus light curves: KIC 6691030, a 3,800 ppm dip every 4.772 days.
NEB sent the signal to a NASA Goddard team that runs a Kepler Bonus search, and they ran it through their own pipeline. Their reply:
"Looks like a great candidate to me!"
"it passed our vetting tests"
"the difference imaging already looks great"
It's a candidate, and a binary could still explain the dip. The code and the data are public, so anyone can check every number: https://t.co/lddoFamvzZ
Miners trained a neural network on NEB. It went through 2,000 faint stars Kepler recorded but never targeted. One signal passed every test: KIC 6691030 dims 0.38% for 1.8 hours every 4.772 days.
Planned for Q1: miners run the search themselves.
https://t.co/i4GiGpHI1f
@__KimMinJae__ Everyone. The chain owns the models, so the weights are public at every block. Watch them live, nothing to install: https://t.co/Ica4dxmA8j
Miners trained this neural network. Each frame is one Kepler training step, over the first 1,384 blocks of NEB.
The neurons are the real weights of the chain. A real signal crosses the network at every step.
Accuracy went from 49.9% to 85.6% on Kepler, 33.7% to 91.8% on TESS.
NEB mainnet, 24 hours and 1,400 blocks in.
The TESS model tells eclipses from junk with 93% accuracy. The Kepler model is at 85% on planet candidates vs. false positives.
Miners trained both. Every mining attempt is one training step.
@sandeep702368 No AMD miner in the official release. The GPU miner needs an NVIDIA GPU with compute capability 7.5 or later. The CPU miner runs on any CPU.
📣NEB mainnet is live.
Every mining attempt is one training step of a small neural network. From block 0 the chain trains two models on NASA data: Kepler transit signals and TESS light curves.
Started November 18, 2026 at 14:00 UTC.
https://t.co/VWVsAepMTM
https://t.co/pG5KjKB65N
NEB 1.0.1 replaces 1.0.0 before the mainnet starts.
Both models now train on NASA exoplanet data: Kepler transit signals and TESS signals. Same start, November 18 at 14:00 UTC.
Installed 1.0.0? Reinstall.
https://t.co/YUQIN13D3I
The telescope's set up. The sky's clear.
For years, a space telescope named Kepler watched the same patch of sky and logged every flicker of its stars. Some flickers are noise. Some are planets crossing in front of their sun.
NEB's miners teach a small network to tell them apart, one block at a time. The model lives on the chain, and anyone can replay how it learned.
Mainnet starts November 18, 14:00 UTC.
📢 NEB 1.0.0 is out, the mainnet release.
The mainnet starts on Wednesday, November 18, at 14:00 UTC. Nodes and miners can be started now: the node shows the time left, the miner waits for its first job.
The genesis block pays 215 addresses 10 % of what they held on the two testnets. For the second testnet the balances were read at block 12900.
neb-miner-cpu has a new engine, --engine table. On one CPU thread it runs 70x the reference on the Kepler network and 213x on the retina network, with the same tickets.
Release: https://t.co/hLAkqoAWuV
Site: https://t.co/VWVsAepMTM