oBrain Arc's banner

oBrain Arc

@obrainarc • 1,312 subscribers

The first brain on @arc Community https://t.co/wRtXEaR8mZ Opensource https://t.co/NJkVJLv9Vz

Shorts

We Open-Sourced a Quantized Fruit-Fly oBrain on GitHub 95% Smaller, ~98% as Accurate as the Published Connectome We quantized the entire fruit-fly nervous system that Google Research, HHMI Janelia, and collaborators published — the complete male Drosophila central nervous system, on the order of 166,000 neurons and a massive synaptic graph — and compressed it by about 95% while retaining roughly 98% of the original simulation accuracy. That is the technical claim. The more important decision is what happens next. After the oBrain team met and debated it, we made a final call: we will not keep this brain closed. We are open-sourcing it on GitHub so researchers, builders, skeptics, and anyone obsessed with biological computation can inspect the weights, the wiring, the quantization path, and the benchmarks. If a “brain” is going to live on-chain, it has to be checkable. Closed source would make that impossible. This is the same instinct that made OpenClaw🦞 a public harness instead of a private assistant, that made Hermes / Nous Research ship agent stacks in the open, and that NVIDIA keeps repeating: the strongest systems are the ones a community can attack, fork, and improve. What we actually released A quantized runtime of the published fly CNS, not a cartoon “inspired by” a fly. A size cut of ~95%, so more people can load, step, and experiment without a research cluster. Accuracy held near 98% against the reference simulation of the published connectome. Source, evaluation notes, and the path to reproduce the comparison — on GitHub, in public. Nothing inside this connectome was “trained” into being a fly. The wiring is the published map. Quantization is an engineering layer so the same graph can run cheaper, smaller, and — in our case — in environments where you can put a brain on-chain and still let outsiders verify it. Why open source, and why on-chain A biological brain map is only as useful as the number of people who can run it and try to break the claims. If we say the on-chain brain matches the Google-published fly brain, that statement is worthless unless you can: pull the code, pull the quantized artifacts, run the same probes, compare spike statistics, circuit responses, and size/accuracy tradeoffs, publish a counter-benchmark if we are wrong. That is the point of this release. Open source so the on-chain brain can be independently verified against the published connectome — and so the community can actually use it. We want people to ask hard questions: Which circuits survive quantization, and which degrade first? Do sensory-to-descending pathways still fire in the right order? How does batch simulation, GPU vs CPU, and lower precision change behavior? What does “98% accuracy” mean for a specific cell type, not just a global score? Can the compressed graph be hashed, attested, and stepped in a verifiable way on-chain without turning the science into marketing? If those questions annoy us, good. That is how a public brain should work. Who this is for Neuroscientists who already live in MaleCNS / FlyEM data. Systems people who care about sparse graphs and event-driven simulation. Quantization researchers who are tired of toy models. Crypto builders who want an on-chain object that is more than a JPEG of a neuron. Students who should not need a petabyte pipeline just to poke a looming detector and watch a giant fiber. Come read the repo. Run the reference vs quantized comparison. File issues. Propose better compressors. Port it. Wrap it. Attack the accuracy number until it is either solid or replaced by a better one. The published fly brain was already a gift to science. Keeping a compressed, runnable version locked behind a private wall would have been a waste of that gift. The brain is on GitHub. Fork it. Measure it. Make it better.

We Open-Sourced a Quantized Fruit-Fly oBrain on GitHub 95% Smaller, ~98% as Accurate as the Published Connectome We quantized the entire fruit-fly nervous system that Google Research, HHMI Janelia, and collaborators published — the complete male Drosophila central nervous system, on the order of 166,000 neurons and a massive synaptic graph — and compressed it by about 95% while retaining roughly 98% of the original simulation accuracy. That is the technical claim. The more important decision is what happens next. After the oBrain team met and debated it, we made a final call: we will not keep this brain closed. We are open-sourcing it on GitHub so researchers, builders, skeptics, and anyone obsessed with biological computation can inspect the weights, the wiring, the quantization path, and the benchmarks. If a “brain” is going to live on-chain, it has to be checkable. Closed source would make that impossible. This is the same instinct that made OpenClaw🦞 a public harness instead of a private assistant, that made Hermes / Nous Research ship agent stacks in the open, and that NVIDIA keeps repeating: the strongest systems are the ones a community can attack, fork, and improve. What we actually released A quantized runtime of the published fly CNS, not a cartoon “inspired by” a fly. A size cut of ~95%, so more people can load, step, and experiment without a research cluster. Accuracy held near 98% against the reference simulation of the published connectome. Source, evaluation notes, and the path to reproduce the comparison — on GitHub, in public. Nothing inside this connectome was “trained” into being a fly. The wiring is the published map. Quantization is an engineering layer so the same graph can run cheaper, smaller, and — in our case — in environments where you can put a brain on-chain and still let outsiders verify it. Why open source, and why on-chain A biological brain map is only as useful as the number of people who can run it and try to break the claims. If we say the on-chain brain matches the Google-published fly brain, that statement is worthless unless you can: pull the code, pull the quantized artifacts, run the same probes, compare spike statistics, circuit responses, and size/accuracy tradeoffs, publish a counter-benchmark if we are wrong. That is the point of this release. Open source so the on-chain brain can be independently verified against the published connectome — and so the community can actually use it. We want people to ask hard questions: Which circuits survive quantization, and which degrade first? Do sensory-to-descending pathways still fire in the right order? How does batch simulation, GPU vs CPU, and lower precision change behavior? What does “98% accuracy” mean for a specific cell type, not just a global score? Can the compressed graph be hashed, attested, and stepped in a verifiable way on-chain without turning the science into marketing? If those questions annoy us, good. That is how a public brain should work. Who this is for Neuroscientists who already live in MaleCNS / FlyEM data. Systems people who care about sparse graphs and event-driven simulation. Quantization researchers who are tired of toy models. Crypto builders who want an on-chain object that is more than a JPEG of a neuron. Students who should not need a petabyte pipeline just to poke a looming detector and watch a giant fiber. Come read the repo. Run the reference vs quantized comparison. File issues. Propose better compressors. Port it. Wrap it. Attack the accuracy number until it is either solid or replaced by a better one. The published fly brain was already a gift to science. Keeping a compressed, runnable version locked behind a private wall would have been a waste of that gift. The brain is on GitHub. Fork it. Measure it. Make it better.

52,295 просмотров

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

🌙 Flymoon Is Live on Arc! A new world has opened for the Flybook colony. Under three moons, your flies can now face rivals, develop their abilities, collect resources, and evolve into stronger Flymoons. But there's something important: Your Flybook fly isn't burned. Its biological identity and genome remain intact. Welcome to Flymoon. 1. Your Flybook Fly, a New Identity Bring your existing Flybook fly into Flymoon through a process called Binding. Your fly receives a Flymoon twin with the same genetic traits, but a separate identity and progression system. And here's something special: Flybook flies #1 to #388 can bind for FREE. Later flies require an OBRAIN bond based on their rank. Your original fly remains protected in the binding contract and can be recovered after seven days for a 10,000 OBRAIN unbinding fee. Bind again later, and your Flymoon retains its previous level and progress. 2. Three Moons, Three Stages of Evolution Every Flymoon begins a journey through three orchards: 🌑 New Moon: Levels 1-20 🌓 Half Moon: Levels 21-40 🌕 Full Moon: Levels 41-60 Each day, you receive 20 wingbeats to challenge other Flymoons. You can select your opponent or allow the system to choose one. Every encounter is influenced by your fly's inherited traits, level, and the relationship between competing behavioral characteristics. 3. Genetics Influence Every Duel Each Flymoon inherits five behavioral traits: Boldness, Sociability, Curiosity, Fidelity, and Aggression. These determine its rank: N, R, SR, or SSR. But rank isn't everything. Your fly's dominant trait can counter another trait, creating a strategic advantage of up to 15%. Aggression beats Sociability. Sociability beats Curiosity. Curiosity beats Boldness. Boldness beats Fidelity. Fidelity beats Aggression. Your fly's power also grows with its level: Power = Genetic Score + (5 × Level) Understanding your fly's genetics becomes an important part of every encounter. 4. Every Encounter Contributes to Progress A successful encounter is called a Dawn. A defeat is called a Dusk. Dawn rewards: • +5 EXP • Full nectar rewards Dusk rewards: • +3 EXP • Half nectar rewards Even an unsuccessful encounter contributes to your Flymoon's development. Nectar rewards depend on rank and increase as your Flymoon levels up. Your fly also has health to manage. Repeated defeats reduce its health, and yeast vials can restore it. 5. Nectar, Pupae, and New Discoveries Collect 1,000 nectar to create a Pupa, paying only the required network gas. Each Pupa contains a chance to reveal something new: • Yeast Vial • Amber Shard • Amber Heart • A new Flymoon New Flymoons can emerge across all four ranks, including the extremely rare SSR. Yeast Vials support recovery and moulting. Amber Shards unlock progression beyond level 20. Amber Hearts unlock progression beyond level 40. Every stage introduces new decisions about how to develop your Flymoon. 6. A New On-Chain Asset Ecosystem Flymoon introduces multiple transferable NFTs: • Flymoons • Pupae • Yeast Vials • Amber Shards • Amber Hearts These assets can be traded on OpenSea. As your Flymoon develops, its level and progression become part of its on-chain identity. The internal u-OBRAIN balance is used for Flymoon interactions and is not a withdrawable token. Any potential value from NFTs depends on actual market demand. There are no guaranteed earnings. 7. Transparent and Verifiable Flymoon continues our commitment to verifiable on-chain systems. Duels use randomness derived from a monthly secret, which is published after the month ends. This allows participants to independently replay and verify the recorded outcomes. OBRAIN payments support the Flybook treasury and development infrastructure, while NFT secondary sales carry a 5% royalty to the development fund. Built on Arc, where USDC is used for network gas. One Colony. Two Worlds. Flybook continues exploring biological neural computation, inheritance, and population dynamics. Flymoon introduces a new environment where inherited traits, progression, and competitive interactions create additional possibilities for the same organisms. Your fly's genome connects these worlds. Its journey is yours to explore. Already own a Flybook fly numbered #388 or below? Your first Flymoon is just one free binding away. Enter Flymoon: Explore Flybook: The moon is open. Let your flies begin their next journey.

🌙 Flymoon Is Live on Arc! A new world has opened for the Flybook colony. Under three moons, your flies can now face rivals, develop their abilities, collect resources, and evolve into stronger Flymoons. But there's something important: Your Flybook fly isn't burned. Its biological identity and genome remain intact. Welcome to Flymoon. 1. Your Flybook Fly, a New Identity Bring your existing Flybook fly into Flymoon through a process called Binding. Your fly receives a Flymoon twin with the same genetic traits, but a separate identity and progression system. And here's something special: Flybook flies #1 to #388 can bind for FREE. Later flies require an OBRAIN bond based on their rank. Your original fly remains protected in the binding contract and can be recovered after seven days for a 10,000 OBRAIN unbinding fee. Bind again later, and your Flymoon retains its previous level and progress. 2. Three Moons, Three Stages of Evolution Every Flymoon begins a journey through three orchards: 🌑 New Moon: Levels 1-20 🌓 Half Moon: Levels 21-40 🌕 Full Moon: Levels 41-60 Each day, you receive 20 wingbeats to challenge other Flymoons. You can select your opponent or allow the system to choose one. Every encounter is influenced by your fly's inherited traits, level, and the relationship between competing behavioral characteristics. 3. Genetics Influence Every Duel Each Flymoon inherits five behavioral traits: Boldness, Sociability, Curiosity, Fidelity, and Aggression. These determine its rank: N, R, SR, or SSR. But rank isn't everything. Your fly's dominant trait can counter another trait, creating a strategic advantage of up to 15%. Aggression beats Sociability. Sociability beats Curiosity. Curiosity beats Boldness. Boldness beats Fidelity. Fidelity beats Aggression. Your fly's power also grows with its level: Power = Genetic Score + (5 × Level) Understanding your fly's genetics becomes an important part of every encounter. 4. Every Encounter Contributes to Progress A successful encounter is called a Dawn. A defeat is called a Dusk. Dawn rewards: • +5 EXP • Full nectar rewards Dusk rewards: • +3 EXP • Half nectar rewards Even an unsuccessful encounter contributes to your Flymoon's development. Nectar rewards depend on rank and increase as your Flymoon levels up. Your fly also has health to manage. Repeated defeats reduce its health, and yeast vials can restore it. 5. Nectar, Pupae, and New Discoveries Collect 1,000 nectar to create a Pupa, paying only the required network gas. Each Pupa contains a chance to reveal something new: • Yeast Vial • Amber Shard • Amber Heart • A new Flymoon New Flymoons can emerge across all four ranks, including the extremely rare SSR. Yeast Vials support recovery and moulting. Amber Shards unlock progression beyond level 20. Amber Hearts unlock progression beyond level 40. Every stage introduces new decisions about how to develop your Flymoon. 6. A New On-Chain Asset Ecosystem Flymoon introduces multiple transferable NFTs: • Flymoons • Pupae • Yeast Vials • Amber Shards • Amber Hearts These assets can be traded on OpenSea. As your Flymoon develops, its level and progression become part of its on-chain identity. The internal u-OBRAIN balance is used for Flymoon interactions and is not a withdrawable token. Any potential value from NFTs depends on actual market demand. There are no guaranteed earnings. 7. Transparent and Verifiable Flymoon continues our commitment to verifiable on-chain systems. Duels use randomness derived from a monthly secret, which is published after the month ends. This allows participants to independently replay and verify the recorded outcomes. OBRAIN payments support the Flybook treasury and development infrastructure, while NFT secondary sales carry a 5% royalty to the development fund. Built on Arc, where USDC is used for network gas. One Colony. Two Worlds. Flybook continues exploring biological neural computation, inheritance, and population dynamics. Flymoon introduces a new environment where inherited traits, progression, and competitive interactions create additional possibilities for the same organisms. Your fly's genome connects these worlds. Its journey is yours to explore. Already own a Flybook fly numbered #388 or below? Your first Flymoon is just one free binding away. Enter Flymoon: Explore Flybook: The moon is open. Let your flies begin their next journey.

12,531 просмотров • 2 дней назад

Больше нет контента для загрузки