Archive
22 artifacts published in this period.
How well can an adversary reconstruct your training data from the gradients you published? Drag the batch size; toggle DP budget ε. The curve shows estimated reconstruction quality (PSNR) from Geiping et al. — and the tradeoff where DP noise finally closes the window at the cost of model utility.
How quickly each Foundry fork property diverges from mainnet reality. Five divergences — blockhash entropy, MEV competition, basefee, TWAP, and oracle price — rated by blocks until first silent failure. Click a bar to inspect fork behavior vs. production.
Three interactive attack surfaces on an ElizaOS-style Web3 AI agent context pipeline — prompt injection, memory injection, and data feed manipulation — with ASR figures from CrAIBench.
Six commitment schemes plotted on proof size vs. trust model: KZG (EIP-4844 blobs), IPA/Pedersen (Verkle state), Groth16, PLONK, FRI-STARK, and Merkle Patricia Trie. Click any dot to inspect.
Step through a CoW Protocol batch auction: intents arrive, the batch seals, solvers compete, settlement executes on-chain, and surplus flows back to traders.
Step through a DeFi AI agent's 8-step execution cycle as a TrajAD-style detector scores each step in real time. Inject a fault to watch a poisoned decision spike to 0.97 and trigger a bisection dispute game that isolates the bad step in 3 rounds.
Interactive ADL priority queue: drag the shortfall slider past the insurance fund to trigger auto-deleveraging. Positions close highest-profit-% first.
Interactive calculator: watch how deposit-borrow loops compound leverage toward the 1/(1−LTV) asymptote. Adjust LTV and loop count to see effective leverage, the liquidation trigger, and the loss from a single price move.
Adjust memory entries per call, call frequency, and gas price to see how monthly on-chain cost diverges across four persistent-memory architectures: SSTORE, event logs, IPFS+CID, and TEE.
Interactive chart of Aave V3's USDC two-slope interest rate model on Base. Drag the slider to explore how borrow and supply APYs respond to utilization — and see the sharp kink at 90% that makes over-borrowing expensive.