Cerebr 入门套件产出链上神经元电路Cerebr Starter Kits Yield Onchain Neuron Circuits
Cerebr Genesis 入门套件已被多位建造者领取并用于流片神经元。免费 NAND 被转化为在 X Layer 上链上运行的真实电路，其中包括一枚 16 门 threshold neuron，它用尽了套件内全部晶体管。此前 Cerebr 神经处理器一日新增 18 套入门套件，本次进展显示套件已开始产生链上电路实例。The Cerebr Genesis starter kits have been claimed by multiple builders and used to tape out neurons. Free NAND was turned into real circuits running onchain on X Layer, including a 16-gate threshold neuron that used every last transistor in its kit. This follows the earlier one-day addition of 18 Cerebr starter kits, showing the kits are now producing onchain circuit instances.
来源 · 6 条Sources · 6
- @NetLayerLabs
Introducing Cerebr: a neural processor, taped out onchain We built a compiler that turns neurons into NAND gates and tapes them out on X Layer via TapeOut. Every transistor is a synapse; every circuit is a neuron anyone can run with eval(). → Train a network in your browser and tape it out → Play
- @encrypt_wizard
We just submitted Cerebr to the TapeOut Genesis Transistor Hackathon, and I'd love for you to break it. Cerebr compiles neurons into NAND gates and tapes them out on @XLayerOfficial, so every circuit is a neuron anyone can run onchain with eval(). Try to beat the Arena bot: a 590-gate neural net
- @NetLayerLabs
Cerebr is in for the TapeOut Genesis Transistor Hackathon by @Ignixbot, live on @XLayerOfficial and built on @TapeOutWorld in a day, builders claimed 18 starter kits and taped out their own neurons, taking the processor from 17 to 28 circuits, each one a real NAND netlist anyone can run onchain wi
- @NetLayerLabs
We're NetLayer Labs, a small team based in Europe building onchain infrastructure Cerebr is our entry for the TapeOut Genesis Transistor Hackathon: a neural processor on X Layer that compiles neurons into NAND gates, so every circuit is a neuron anyone can run onchain. Everything is open source
- @boostbob
TapeOut 为每个处理器提供两个合约：可铸造的晶体管，以及通过销毁晶体管流片得到的电路。Cerebr 补上了缺失的一环：一个把神经网络编译成这些电路的编译器。 Cerebr 将神经元编译为真实的 NAND 网表，并在自己的 TapeOut 处理器上流片。每个晶体管都是一个突触，每个电路都是一个归你所有的神经元，任何人都可以通过 eval() 在链上运行推理。 目前 Cerebr 原生支持的是二值输入、二值输出的阈值神经网络，但权重可以是整数，不局限于 0/1 或 ±1。 运行常见的浮点神经网络，还需要实现乘加、激活函数等电路，门数和链上执行成本会大幅增加。 它当前展示的是小型、离散
- @encrypt_wizard
Huge thanks to every builder who grabbed a Genesis starter kit and taped out a neuron of their own on Cerebr 🙌 you turned free NAND into real circuits running onchain on X Layer, including a 16-gate threshold neuron that used every last transistor in its kit. And to everyone who took on the Aren