标签：#EU

Training a network from scratch in raw NumPy, quantizing it to int8, and running it as ~80 lines of dependency-free JavaScript — with a parity test proving the browser matches Python to 1e-6. Why bother? MNIST is a solved problem Digit recognition is the "hello world" of ML — that's exactly why I used it. The model isn't the point. The point is everything around the model, which happens to be the part that matters in production work too: training without a framework, compressing for deployment, running inference in a constrained environment, and proving the deployed system matches the trained one. Training: just NumPy and math The network is a 784→128→64→10 MLP — hand-written forward pass, backpropagation, and Adam optimizer. No autograd, no framework: # backward pass, by hand dz3 = ( probs - y_batch ) / batch_size grads_w [ 2 ] = a2 . T @ dz3 da2 = dz3 @ weights [ 2 ]. T dz2 = da2 * ( z2 > 0 ) # ReLU mask grads_w [ 1 ] = a1 . T @ dz2 ... One trick that matters for a drawing demo specifically: shift augmentation . MNIST digits are centered; humans draw wherever they like. Training on randomly translated copies makes the model tolerant of sloppy placement. Combined with MNIST-style preprocessing at inference (crop to bounding box, scale into a 20×20 box, center by center-of-mass), real-world doodles classify reliably. Final test accuracy: 98.2% . Compression: int8 in 15 lines A float32 weight file would be ~430 KB. Symmetric int8 quantization cuts it ~4×: scale = np . abs ( w ). max () / 127.0 q = np . clip ( np . round ( w / scale ), - 127 , 127 ). astype ( np . int8 ) One scale factor per layer, weights stored as base64 in JSON: 145 KB total , and quantized test accuracy is identical to float — 98.2%. Inference: ~80 lines of plain JavaScript In the browser, the weights are dequantized once on load, and inference is three matrix-vector products with ReLU and a softmax. ~109K multiply-adds — about a microsecond-scale problem for any modern device. No TensorFlow.js (t

We take the Falcon 9 rocket for granted. But we probably shouldn't.

Every week I see the same question in AI governance communities: "We already have NIST AI RMF implemented. Does that cover our EU AI Act obligations?" The honest answer is: sometimes yes, sometimes partially, and sometimes not at all. The problem is that nobody had built a clean, free, interactive tool that showed exactly which controls map to which, how strong those mappings actually are, and where the genuine gaps are. So I built one. Live tool: suhanasayyad.github.io GitHub: SuhanaSayyad / eu-ai-act-crosswalk-tool Interactive crosswalk mapping EU AI Act obligations to NIST AI RMF and ISO 42001 controls, with mapping strength indicators, gap analysis, and source links. 30 controls mapped. Free and open source. EU AI Act × NIST AI RMF × ISO 42001 - Interactive Compliance Crosswalk Tool An open-source tool that maps EU AI Act obligations to their equivalents in NIST AI RMF and ISO 42001, with mapping strength indicators, gap analysis, and source document links. Built for compliance teams, AI governance practitioners, and anyone trying to understand how these three frameworks relate to each other. Live demo: https://suhanasayyad.github.io/eu-ai-act-crosswalk-tool Built by: Suhana Sayyad | MSc Cybersecurity, TUS Athlone Why I built this Every organisation dealing with the EU AI Act is being asked the same questions: "We already have NIST AI RMF controls in place. Does that cover our EU AI Act obligations?" "We're pursuing ISO 42001 certification. Does that satisfy the regulation?" The honest answer is: sometimes yes, sometimes partially, and sometimes not at all. The problem is that nobody had built a clean, free, interactive tool that showed exactly which… View on GitHub What the tool does The EU AI Act / NIST AI RMF / ISO 42001 Interactive Crosswalk Tool maps 30 EU AI Act obligations to their nearest equivalents in NIST AI RMF and ISO 42001. For each mapping it shows a strength rating - Strong, Partial, Indirect, or No Equivalent - so compliance teams know which map

Welcome to our weekly digest, where we unpack the latest in account and chain abstraction and the broader infrastructure shaping Ethereum. This week: Base ships its first independent upgrade with a TEE+ZK multiproof system and confirms native AA is next; Biconomy turns the ERC-8211 smart batching standard into a TypeScript SDK; LI.FI launches an enterprise intents engine for stablecoin and RWA flows; and Vitalik steps back from technical essays to write fiction. Base Launches Azul, Bringing Multiproofs to Coinbase's L2 Biconomy Ships Smart Batching SDK for ERC-8211 LI.FI Launches Intents Engine for Enterprise Cross-Chain Flows Vitalik Pivots to Fiction and Floats a "Trust Dependency" Framework Please fasten your belts! Base Launches Azul, Bringing Multiproofs to Coinbase’s L2 Base activated Azul on mainnet on May 28, its first network upgrade built entirely on its own stack. The headline feature is a multiproof system that pairs Trusted Execution Environment (TEE) proofs with Zero-Knowledge (ZK) proofs, advancing the Coinbase-incubated L2 toward Stage 2 decentralization. Either proof type can finalize a withdrawal independently, but when both agree, finality drops to as little as one day, far faster than the typical multi-day optimistic rollup wait. Crucially, permissionless ZK proofs can override permissioned TEE proofs if the two conflict, a design Base says meaningfully improves censorship resistance. The upgrade also makes base reth the sole execution client and introduces a new consensus client, phasing out older software. Node operators must migrate to the new stack to stay in sync. For AA and chain abstraction builders, the more important signal is what comes next: Base confirmed its end-of-June upgrade will include native account abstraction, an enshrined token standard and Flashblock Access Lists. The largest L2 by activity moving toward native AA is a meaningful pull on the whole ecosystem. Biconomy Ships Smart Batching SDK for ERC-8211 Biconomy released t

This open-source community project lets you create a StumbleUpon-like experience for recommending your favorite sites.

🏦 Day 6 of 7: Building a Mini Uniswap in 80 Lines of Solidity Imagine a vending machine. It has 1,000 coffee beans and 1,000 coins. No menu, no cashier — just one iron rule: the product of the two numbers inside must never decrease. That's it! This is how Uniswap works — and this is what I built on Day 6, coming from .NET. Here's how, why it's elegant, and where you can step on a rake. Why an Order Book Doesn't Work on a Blockchain Traditional exchanges — Binance, NYSE, any CEX — run on an order book . Market makers post bids and asks. A matching engine pairs them. Millions of updates per second, all in a centralised database. In a blockchain, this is impossible. Transactions take 12 seconds. Every state change costs gas. Storing millions of constantly changing orders would eat all the profit before a single trade completes. Uniswap's solution: replace the order book with a liquidity pool — a smart contract holding two tokens — and replace the matching engine with pure math. Just a formula — below. x · y = k — The Formula That Broke Finance The Constant Product Invariant : x · y = k Where x is the reserve of Token0, y is the reserve of Token1, and k is a constant that must never decrease during swaps. When a trader sells Token0 into the pool, x increases. To keep k constant, y must decrease — the contract sends out Token1. The price is determined automatically by the ratio of reserves. Live example with numbers: Pool: 1,000 Token0, 1,000 Token1. k = 1,000,000. Trader sells 100 Token0: amountOut = (reserveOut × amountIn) / (reserveIn + amountIn) amountOut = (1000 × 100) / (1000 + 100) amountOut = 100,000 / 1,100 amountOut ≈ 90.9 Token1 The trader gets ~90.9, not 100. That gap is slippage — and it's not a bug. It's the formula protecting the pool. The more you buy relative to pool size, the worse your price gets. Naturally. Mathematically. After the swap: pool has 1,100 Token0 and ~909.1 Token1. k ≈ 1,000,000. Invariant holds. The Contract: SimpleAMM Three functions.

A complete ML pipeline: engine, backprop, binary format, and a live browser demo. Zero dependencies. Under 200 KB total. If you have built machine-learning projects before, you have probably done it by importing PyTorch, TensorFlow, or scikit-learn and calling .fit() . Those are excellent libraries. This article is about what happens when you deliberately do not use them — when you build every piece of the pipeline yourself, in a language that compiles to WebAssembly, and the result runs live in the browser with no server, no Python, and no cloud bill. Here is the live demo: move four sliders, watch the predicted Iris species update in real time. The model is running entirely inside your browser tab, loaded from a 1.1 KB binary file, powered by ~100 KB of WebAssembly compiled from pure Rust. This is the story of how I built it and why the engineering choices made it work. Why Rust? Why WebAssembly? Why zero dependencies? Three constraints drove every design decision. WASM requires no_std or a carefully limited std . The wasm32-unknown-unknown target has no operating system, no file system, and no libc. A crate that links against rand , ndarray , or any library that makes OS calls will not compile to it without significant plumbing. An engine built from nothing but the Rust standard library compiles cleanly to every target, including WASM. A zero-dependency std -only crate is uniquely auditable. There are no transitive dependency trees to vet, no supply-chain risks, no version conflicts. Every line of code that runs in the user's browser lives in this repository. The deployment story becomes the technical story. A 100 KB WASM blob that runs locally in the browser is not just a cost optimisation — it is a privacy guarantee (user inputs never leave the machine) and a latency guarantee (inference is microseconds, not a round trip to a cloud API). That story is only possible because the engine has no external dependencies that would bloat the binary. The architecture: ei

Every line of your smart contract costs something. Some lines cost more than others. In this part of our gas saving series, we’ll explore how to write smarter Solidity code that keeps your contract lean and efficient. Here are six simple and practical ways to reduce gas costs while writing Solidity smart contracts. 1. Use payable Only When Needed, But Know It Saves Gas In Solidity, a function marked payable can actually use slightly less gas than a non-payable one. Even if you're not sending ETH, the EVM skips some internal checks when the function is marked payable. See this example: function hello() external payable {} // 21,137 gas function hello2() external {} // 21,161 gas That tiny difference may not seem like much, but across thousands of calls, it adds up. Only use payable when your function is actually meant to accept ETH 2. Use unchecked for Safe Arithmetic When You’re Sure Since Solidity 0.8.0, all arithmetic operations automatically check for overflows and underflows. While this makes contracts safer, it also uses extra gas. When you're certain that overflow won't occur, you can use the unchecked keyword to skip these safety checks. uint256 public myNumber = 0; function increment() external { unchecked { myNumber++; } } Gas used: 24,347 (much cheaper than using safe math) Warning: Use unchecked carefully. Only when you're confident there's no risk of overflow. 3. Turn On the Solidity Optimizer The Solidity Optimizer is like a smart helper that cleans up and tightens your compiled bytecode. It does not change how your contract works, but it removes waste and makes it cheaper to run. If you’re using tools like Hardhat or Remix, always enable the Optimizer before deploying to mainnet. 4. Use uint256 Instead of Smaller Integers (Most of the Time) Smaller types like uint8 or uint16 might look more efficient, but they can cost more gas during execution. That’s because the EVM automatically converts them to uint256 behind the scenes. So, if you're not tightly p