Maximum and minimum causal effects of physical processes

Kaumudibikash Goswami,Giulio Chiribella
2024-05-03
Abstract:We introduce two quantitative measures of the strength of causal relations in general physical theories. These two measures called the maximum and minimum causal effect, capture the maximum and minimum changes in a physical system induced by changes in another system. In quantum theory, we show that both measures possess important properties, such as continuity and faithfulness, and can be evaluated through optimization over orthogonal pairs of input states. For the maximum causal effect, we provide numerical lower bounds based on a variational algorithm, which can be used to estimate the strength of causal relations without performing a full quantum process tomography. To illustrate our algorithm, we analyze two paradigmatic examples, the first involving a coherent quantum superposition of direct cause and common cause, and the second involving communication through a coherent quantum superposition of two completely depolarizing channels.
Quantum Physics
What problem does this paper attempt to address?
The paper mainly discusses the quantification of causal effects in physical processes. The authors propose two new measures of causality strength, called Maximum Causal Effect (GCE max) and Minimum Causal Effect (GCE min). These two metrics are applicable to general physical theories and are inspired by the Average Causal Effect (ACE) in classical theory. In quantum theory, these two measures possess continuity and fidelity, and can be evaluated by optimizing orthogonal input states. GCE max is defined as the maximum distance under the output state, while GCE min is defined as the minimum ratio between the distance between the output state and the corresponding input state. They respectively represent the best and worst cases of causality between input and output in the process. For quantum systems, these two quantities are related to the known quantum ACE quantity, and GCE min can serve as a lower bound for the classical capacity of quantum channels. In the paper, the authors also develop a variational algorithm to estimate the Maximum Causal Effect (QCEmax) of quantum channels without performing complete quantum process tomography, which is infeasible for large-scale quantum systems. The effectiveness of the algorithm is demonstrated through two examples, including a quantum superposition state involving common causes and direct causes, and a coherent superposition involving quantum channels where individual channels have no causal effect. In summary, the aim of the paper is to provide new tools for quantifying quantum causal relationships and proposes a practical method to compute the strength of these relationships.