Researchers Tao Wang and Yun Shang at Chinese Academy of Sciences, have developed a new error mitigation technique to improve the reliability of calculations performed on contemporary quantum computers. Their work introduces a hybrid Gaussian-exponential extrapolation scheme specifically tailored for quantum circuits exhibiting periodic structure, a common characteristic of many quantum algorithms. By modelling noise amplification using a log-normal distribution and incorporating Gaussian variance corrections, the method requires minimal prior noise characterisation and demonstrably reduces bias in calculations across a range of circuit types, including Trotterized Ising dynamics, random circuits, and Grover search. This advancement represents a key step towards obtaining more accurate results from near-term quantum hardware, which is currently limited by the inherent fragility of quantum states and susceptibility to environmental disturbances. Log-normal noise modelling diminishes bias in deep periodic quantum circuits Up to 30% bias reduction in calculations was achieved through this new method, compared to previous extrapolation variants. Existing error mitigation techniques often struggle to maintain accuracy with increasingly complex and deep circuits. These circuits involve many quantum operations, previously rendering accurate results unattainable without substantial qubit overhead, which increases the resources required for computation. Dr. Patrick Draper and Dr. Pranav Patel conceived this new technique for quantum circuits with periodic structure, circuits constructed from repeating blocks of operations. The core innovation lies in modelling noise amplification as a log-normal distribution, a statistically robust approach grounded in the observation that errors tend to accumulate multiplicatively rather than additively. This multiplicative accumulation leads to a distribution of errors that closely resembles a log-normal distribution, allowing for more accurate extrapolation to the zero-noise limit. The hybrid model accurately characterises noise behaviour without requiring prior knowledge of the specific noise affecting the quantum computer, significantly simplifying implementation and broadening its applicability to diverse hardware platforms. This is a crucial