Archive
22 articles published in this period.
ElizaOS agents treat their memory store as ground truth — their own past. CrAIBench tests 685 attack cases across 33 Web3 action types and finds memory injection hits 55.1% ASR even on Claude Sonnet 3.7, while every off-the-shelf detector fails.
On-chain AI agents reset state on every call. Four persistent-memory architectures span a 20× monthly cost gap — SSTORE at $78/month versus IPFS+CID at $3.92. Real June 2026 gas numbers, four trust models, one crossover you won't see coming.
Ethereum's 12-second slot clock creates a hard lower bound on AI agent state freshness. Here's what that means for liquidators, arbitrageurs, and yield rebalancers — and how to engineer around it.
Bitcoin burns ~150 TWh/yr on SHA-256 nonces that prove nothing. Komargodski & Weinstein showed a Freivalds randomisation check turns any matmul into a valid proof-of-work at only 3/(2N) overhead — the same GPU seconds train your model and mine the block.
DeFi interest rate formulas are open-source reward functions. An RL agent reading Aave's two-slope model can compute the optimal adversarial strategy analytically — no learned reward model required.
Most DeFi protocols are upgradeable, and most on-chain AI agents don't know it. Here's how proxy architecture invalidates agent assumptions — and a tiered defense framework to build around it.
A 15× spike on a $17M pool shifts its 30-minute v2 TWAP by 9.3% in one block. The arithmetic-mean accumulator is far weaker than v3's geometric tick accumulator. Covers the manipulation cost formula, the Inverse Finance case study, and minimum safe pool depth.
Chainlink's push oracle only updates when price moves ≥0.5% or an hour passes. That gap costs AI agents real money — and Pyth's pull model closes most of it.
Federated learning keeps raw data local — but the gradients it publishes on-chain are not private. Inversion attacks reconstruct training images from them, and a public ledger makes that window permanent.
Controlling one intermediate pipeline stage is enough to inject a backdoor into a decentralized post-training run — 94% ASR, no data access required. The mechanism, why training loss stays clean, and what it means for networks fine-tuning LLMs across untrusted nodes.