Werracle: Sub-Cent Intra-Block AI Reflex Oracles and Flash-Loan Circuit Breakers for EVM Smart Contracts

Volkan Dağlı; Zerrin Dağlı; Dağhan Dağlı

cs.CRSep 25, 2026

Werracle: Sub-Cent Intra-Block AI Reflex Oracles and Flash-Loan Circuit Breakers for EVM Smart Contracts

Authors: Volkan Dağlı, Zerrin Dağlı, Dağhan Dağlı

Organizations: ITOUCH Bili¸im Sistemleri Ltd. S¸ti., C¸ukurova Teknokent, Adana, Türkiye · Anadolu University, Eski¸sehir, Türkiye · Mersin University, Mersin, Türkiye · Toros Science High School, Mersin, Türkiye

Abstract

Contemporary on-chain artificial intelligence (AI) encounters an intractable Von Neumann memory and latency wall. Storing static floating-point neural weight matrices inside Ethereum Virtual Machine (EVM) storage costs millions of gas, rendering direct on-chain inference impossible. While Zero-Knowledge Machine Learning (ZK-ML) offloads matrix tensor multiplications to off-chain provers, it introduces fatal constraints: 10 to 300 seconds of SNARK proving latency and 250,000 to 500,000 gas per proof verification. Because decentralized finance (DeFi) exploits - such as uncollateralized flash-loan attacks, predatory sandwich MEV, and toxic loss-versus-rebalancing (LVR) flow - occur atomically inside a single block, ZK-ML oracles cannot react in time. Here, we present Werracle, a production-grade, zero-storage on-chain AI decision oracle fitting inside a single 32-byte EVM storage slot (bytes32). Leveraging foundational procedural Mandelbrot escape dynamics (z_{n+1} = z_n^2 + c) established by Dagli et al. (arXiv:2609.25498), Werracle derives continuous non-linear decision hyperplanes from a 24-byte coordinate triplet Theta = (c_x, c_y, zoom). Implemented in pure Solidity bytecode using fixed-point Q16.16 arithmetic (WerrMath.sol), Werracle evaluates a 16-point Pareto micro-grid in only 21,438 gas (under 0.0005 USD on Layer-2 rollups like Base and Arbitrum) with sub-millisecond execution latency. We demonstrate real-world DeFi efficacy via WerracleFeeHook.sol, a Uniswap v4 dynamic swap fee governor that measures orderbook turbulence on-the-fly and atomically adjusts liquidity provider fees between 0.05% and 0.50%. The protocol is formally verified against a 1,000-test cryptographically sealed deterministic verification suite (100.0% pass rate) with telemetry permanently disabled, operating live on a dedicated EVM devnet sandbox (Chain ID 4242).

Figures & tables

Figure 1: End-to-end Werracle on-chain architecture. Incoming transaction metrics (volume, tick velocity, address) are normalized into coordinate shifts. The core engine queries a single 32-byte EVM storage word ( $\Theta=(c_{x},c_{y},\text{zoom})$ ), evaluates Mandelbrot escape dynamics across a 16-point Pareto micro-grid using fixed-point Q16.16 arithmetic ( $\mathtt{WerrMath.sol}$ ), and executes intra-block circuit breaking or fee adjustments in 21,438 gas with sub-millisecond latency.

Execution Phase	EVM Gas	USD (Base L2)
Slot SLOAD & Bit Unpacking	2,342 gas	$0.00004
16-Point Q16.16 Iteration	14,810 gas	$0.00029
Pareto Density Aggregation	2,416 gas	$0.00005
Event Log & ABI Return	1,870 gas	$0.00003
Total Forward Inference	21,438 gas	¡ $0.00042

Table 1: Opcode-Level EVM Gas Profiling

Feature	Web2 Oracle	ZK-ML (EZKL)	Werracle
Model Weights	Gigabytes	Off-chain Prover	0 Bytes
EVM Storage	N/A	Verification Keys	32 Bytes (1 Slot)
Proving Latency	12–36 s	10–300 s	¡ 1 ms (Atomic)
Verification Gas	$\sim$ 60,000 gas	$\sim$ 380,000 gas	21,438 gas
Intra-Block Revert	No	No	Native
Hardware Requirements	Server Node	Dedicated GPU Cluster	None (Pure EVM)

Table 2: Architectural Benchmark Comparison

Figure 2: Architectural benchmark comparisons: (a) On-chain verification gas cost across Web2 oracles, ZK-ML (EZKL), and Werracle. (b) Inference and proving latency in seconds (logarithmic scale). Werracle achieves an 18 $\times$ gas reduction and over 1,000 $\times$ latency reduction compared to ZK-ML proving architectures.

Explore similar work

Zero-Storage Procedural Neural Synthesis via Boundary Dynamics: Formal Verification in Lean 4 and Bare-Metal Gauntlet Validation
Sep 27, 2026Volkan Dağlı, Zerrin Dağlı, Dağhan DağlıClassification Formal Verification
Anadolu University, Türkiye · ITouch Systems, Türkiye · Mersin University, Türkiye +1
PACE: Policy-Attested Contract Execution for Safe AI Agents in Decentralized Finance
Aug 18, 2026Rabimba Karanjai, Yang Lu, Richard Williamson +5AI Agent Safety Runtime Enforcement for AI Agents
Larry · University of Houston, USA · PayPal, USA +1
Jolt Atlas: Verifiable Inference via Lookup Arguments in Zero Knowledge
Feb 19, 2026Wyatt Benno, Alberto Centelles, Antoine Douchet +1Privacy-Preserving ML Memory-Efficient Inference
ICME Labs

Sep 27, 2026J/K move · Enter open · S save

Zero-Storage Procedural Neural Synthesis via Boundary Dynamics: Formal Verification in Lean 4 and Bare-Metal Gauntlet Validation

Volkan Dağlı, Zerrin Dağlı, Dağhan Dağlı

Anadolu University, Türkiye · ITouch Systems, Türkiye · Mersin University, Türkiye+1

Classification Formal Verification

Open paper AskPDF arXiv

Contemporary neural inference architectures rely on dense floating-point weight matrices stored in high-bandwidth memory (VRAM), incurring severe memory-wall bottlenecks and preventing native execution inside deterministic virtual machines like the Ethereum Virtual Machine (EVM). Verifying termination and arithmetic invariants for recursive dynamical systems over continuous domains is generally undecidable in the Blum-Shub-Smale model. Here, we present the formal verification and bare-metal empirical validation of WERR (Waves & Errors) and Phase III Orbital Error Dynamics (OED), a non-tensor decision paradigm that procedurally synthesizes non-linear decision boundaries on demand from a 24-byte coordinate seed

Θ= (c_x, c_y, \text{zoom})

along the boundary of the Mandelbrot set (

\partial\mathcal{M}

). By projecting the recurrence

z_{n+1} = z_n^2 + c

onto the modular residue ring

\mathbb{Z}/9\mathbb{Z}

and the fixed-point domain

\mathbb{Q}_{16.16}

, we establish ten machine-verified theorems in Lean 4 (v4.34.1) with Mathlib4 and zero unproven conjectures (sorry): proving

\mathcal{I}_3 = \{0,3,6\} \subset \mathbb{Z}/9\mathbb{Z}

ideal closure, universal fuel-bounded halting (

\le 9

and

\le 12

steps), absence of

\mathbb{Q}_{16.16}

square overflow below

2^{63}-1

, non-constant boundary escape sensitivity, and a parametric EVM gas bound (

\le 22,557 \le 24,000

gas). Evaluated on a 40-core Dual Intel Xeon server, the vectorized 36-iteration CPU kernel processes 100,000 decisions in 6.49 s (15,397 decisions/s, 0 Bytes VRAM, 15.15x speedup), while a sigmoidal outlier gate suppresses 100.00% of adversarial spikes while preserving 89.60% of clean baseline signals.

Abstract

Explore similar work

Zero-Storage Procedural Neural Synthesis via Boundary Dynamics: Formal Verification in Lean 4 and Bare-Metal Gauntlet Validation

PACE: Policy-Attested Contract Execution for Safe AI Agents in Decentralized Finance

Jolt Atlas: Verifiable Inference via Lookup Arguments in Zero Knowledge